Smart Distribution Mechanisms—Part I: From the Perspectives of Planning

Abstract

To enhance the reliability and resilience of power systems and achieve reliable delivery of power to end users, smart distribution networks (SDNs) play a vital role. The conventional distribution network is transforming into an active one by incorporating a higher degree of automation. Replacing the traditional absence of manual actions, energy delivery is becoming increasingly dependent on intelligent active system management. As an emerging grid modernization concept, the smart grid addresses a wide range of economic and environmental concerns, especially by integrating a wide range of active technologies at distribution level. At the same time, these active technologies are causing a slew of technological problems in terms of power quality and stability. The development of such strategies and approaches that can improve SDN infrastructure in terms of planning, operation, and control has always been essential. As a result, a substantial number of studies have been conducted in these areas over the last 10–15 years. The current literature lacks a combined systematic analysis of the planning, operation, and control of SDN technologies. This paper conducts a systematic survey of the state-of-the-art advancements in SDN planning, operation, and control over the last 10 years. The reviewed literature is structured so that each SDN technology is discussed sequentially from the viewpoints of planning, operation, and then control. A comprehensive analysis of practical SND concepts across the globe is also presented in later sections. The key constraints and future research opportunities in the existing literature are discussed in the final part. This review specifically assists readers in comprehending current trends in SDN planning, operation, and control, as well as identifying the need for further research to contribute to the field.

1. Introduction

An electric power distribution system’s primary obligation is to provide power from substations to end users. However, due to the participation of various stakeholders in the electricity market and environmental concerns, the effective and reliable delivery of power through conventional distribution mechanisms is a serious problem in the current revolution era [1]. Due to deficiency of electrical power from the utility side and in load-shedding cases in rural areas, the concept of distributed generation (DG) integration at distribution level is increasing day by day. However, given the current state of distribution infrastructure, this concept faces certain technical issues. The massive increase in DG unit integration at distribution level would cause substantial planning concerns as well as legal and regulatory disputes [2]. Voltage stability, frequency control, and reserve allocation are all affected by the high intermittency of DGs in a weak distribution network [3]. To tackle these issues and to manage this sporadic integrating nature of DGs, a concept of smart distribution networks (SDNs) is introduced [1,4]. It is a key component of the smart grid, since it provides the major network with a user-oriented source. SDN is a concept of transformation of a passive distribution network to active, because of the accelerated growth of new information and communication technology (ICT), as well as the integration of advanced metering infrastructure (AMI) [5]. Distribution management systems (DMS) and energy management systems (EMS) function as decision support information systems (DSIS) for the synchronization of network remote equipment, while SCADA systems are used to monitor the distribution system. Meeting the UN Sustainable Development Goals (SDGs) and need to accommodate high penetration of DGs, especially RES (solar and wind), in order to reach environmental goals for gas emission reduction and sustainability is the driving force behind the transition of a passive distribution mechanism to active one. Small-scale microgrids (MGs), multimicrogrids (MMGs), virtual power plants (VPPs), smart homes (SHs), smart neighborhoods (SNHs), smart buildings (SBs), and smart cities (SCs) all come under the umbrella of SDNs [1]. Figure 1 depicts the projected profile of market value investment in SGs and SDNs deployment and growth by region from 2017 to 2023.

Thomas Edison pioneered the microgrid concept in 1882 [6]. Small-scale MGs [7] are characterized as building blocks and a subset of SG infrastructure with a unique network protocol. The organization of the MGs is based on the growing penetration of DGs, such as microgenerators, photovoltaic (PV), small-scale wind power plants (WPPs), fuel cells (FC), energy storage systems (ESS), such as solid-oxide batteries and flywheels, supercapacitors, grid-to-vehicle (G2V) and vehicle-to-grid (V2G) technologies and demand response (DR) loads, which provide control capabilities over network operation. Figure 2 According to the US Department of Energy (DOE), the estimation of load supplied by MG technology was about 13,000 MW by the end of 2020 [8]. Figure 3 shows the penetration capacity of MGs in different states of the US by the end of 2020. According to the forecast, the Asia–Pacific region will continue to outperform the rest of the economies in terms of market value until the end of 2023.

MMGs [9] are a unique idea that refers to a high-level structure made up of numerous low voltage (LV) MGs and DG units connected to neighboring medium voltage (MV) feeders at the MV level. MGs, DG units, and MV loads under active DSM control will all be called dynamic units in this type of power system for grid control and management. The ability of some MV receptive loads to receive control requests under a load reduction strategy is seen as a way to ensure additional ancillary services in this new situation. Within the MMG paradigm and the electricity market, the MMGs infrastructure allows for bilateral energy and market price transaction through MGO and MO, respectively [10]. The conceptual diagram of MMG setup is given in Figure 4.

The output power of some unconventional sources, such as PV and WPP, is always intermittent due to variable wind speed and solar irradiance on different time scales throughout the day. The concept of VPPs was introduced to accommodate the integration of intermittent output from DGs, such as PV and WPP. VPP is an aggregation of DG units of various technologies that can operate as a single power plant with the ability to control and maintain the electrical energy flow between these units for better system operation [11]. The key difference between MGs and VPPs is that [12] MGs are capable of both grid-connected and islanding operation. The ability of a microgrid to separate and isolate itself from the utility’s distribution system during brownouts or blackouts is its most convincing feature. A VPP, on the other hand, is not tied to the grid—it always operates in a grid-connected mode. The VPP’s plus point is that it can optimize the whole system without requiring significant infrastructure investments.

Figure 1. Smart grid and SDNs market value by country (2017–2023) (adapted from [13]).

Figure 1. Smart grid and SDNs market value by country (2017–2023) (adapted from [13]).

Figure 2. Recent advances in the smart grid paradigm (adapted from [14]).

Figure 2. Recent advances in the smart grid paradigm (adapted from [14]).

Figure 3. MG penetration capacity in selected states of the US by the end of 2020 (adapted from [15]).

Figure 3. MG penetration capacity in selected states of the US by the end of 2020 (adapted from [15]).

Figure 4. Visionary distribution network for Vision 2030 (adapted from [16]).

Figure 4. Visionary distribution network for Vision 2030 (adapted from [16]).

In the current deregulated energy market, a mechanism is required that can provide insights into energy use and help end users to use energy efficiently while being aware of environmental concerns. According to [17], SH is a concept of integration of various services inside a home using a unified communications network. It ensures the home’s operation is cost-effective, safe, and comfortable, with a high level of smart functionality and versatility. Unlike previous concepts, SH is a smart distribution mechanism (SDM) that mostly relies on the consumption side of the power grid. SH usually consists of sensor networks, HAN, smart information box, home display unit as consumer interface, in-house AC/DC distribution with smart plugs, advanced conversion devices for DG, EV and ESS, and controllable AC/DC loads [18].

SNHs include such features as energy-efficient gadgets, connected gadgets, advanced security solutions and home automation. All of this is intended to make homeowners’ lives easier and give them more control over their homes and energy consumption. The relationships between buildings and the SG have the potential to help manage energy demand in the SG environment. These interactions within buildings in an SNH, as well as the interaction between the SG and the SNH, can be realized using the multiagent system (MAS), where the SNH acts as an MG [19].

SBs refer to any structure that concerns the reliable operation and adopt automated protocols to monitor the building’s operations, such as air conditioning, ventilation, heating, lighting, and security. SBs concept applies to a wide range of loads, including residential, commercial, and industrial [18]. The main objectives regarding adaptation of SB structure are the reduction in GHG emissions, minimizing energy expenses, EV and RES integration and adaptation, and RTP [20]. Consumers can engage in the control, monitoring, and shaping of their energy demands using SB concepts in an SDN environment. SCs point to the need for new urban planning policies.

In the context of SCs, a city is considered smart when developments in human and social capital, as well as conventional (transport) and contemporary ICTs, support long-term economic development and a high quality of life, while also ensuring prudent resource management via participatory governance [21]. The use of IT in SCs would result in a variety of objectives [22], including lower GHG emission like CO₂, improved quality of life, and reduced need for conventional construction developments, as well as improved commercial firms. In a nutshell [18], SCs reveal a decentralized, hierarchical, and independent structure with smart technologies incorporated for solutions to large-scale SDN issues. Advance metering infrastructure (AMI), distributed energy resources (DER), ICTs, and demand-side management (DSM) are prominent technologies for SCs.

There is currently no comprehensive simultaneous review of the planning, operation, and control of SDN mechanisms in the existing literature. Given the relevance of Smart distribution mechanisms (SDMs) and concepts, this paper provides a schematic overview of SDN planning, operation, and control of the last 10 years’ (2010–2021) literature. We provide an overview of recent advances in various approaches and techniques for planning, operating, and control of SDNs. The practical implementation of SDN projects across the globe are also summarized. In the later sections of the paper, key constraints and future research recommendations in the existing literature are provided for each SDN technology. This review could provide an excellent foundation for understanding existing and future developments in planning, operation, and control of SDN mechanisms.

The paper is organized as follows. Section 2 describes the need for active planning, operation, and control and of SDNs. Section 3 describes the reviewed work on the SDMs towards planning, operation, and control perspectives. A summary of practical implementation of SDN projects across the globe is presented in Section 4. Section 5 is a summary of the reviewed work, constraints, and future research recommendations in the existing literature. Finally, the review is concluded in Section 6.

2. Need for Active Planning

With the evaluation and expansion of SDNs under the smart grid umbrella, the conventional distribution network planning approach must be revisited from a new perspective, taking into account the infrastructure complexities of emerging technologies (such as deregulated electricity markets, solar and wind power), as well as their correlations [23]. In a traditional distribution grid, distribution network planning ensures feasible economic solutions and the sufficient substation and feeder capability that is available to fulfil load projections during the planning horizon [24]. However, the traditional distribution system planning becomes further complicated when DGs get involved at distribution level. Several technical issues, such as energy market, power efficiency and reliability, cost and risk control, and load uncertainties, are posed as a result of the inclusion of DGs at the distribution level.

The concept of SDNs is supposed to be compliant enough with intelligent technology to support a wide range of DG types, including RESs, different ESS, EVs, and DR. Furthermore, major power sector players are being urged to consider the need to resolve difficult problems and restructure the distribution network in accordance with the SG environment [18]. The key objectives of different SDN planning approaches are: (i) minimization of new facility installation costs (i.e., substations, feeders, and feeder branches), (ii) minimization of capacity upgrading of the existing facilities, (iii) cost of the operation (maintenance and electricity loss) minimization, and (iv) maximization of system reliability. These objectives must take into account network limitations such as substation and feeder power, as well as the overall permissible node voltage divergence [25]. The classical and desired planning approaches for SDNs are presented in Figure 5. The brief hierarchical framework for carried work is presented in Figure 6.

3. Smart Distribution Mechanisms

3.1. Active Radial Distribution Networks

When DGs are connected to a passive distribution network, the network is termed an active distribution network. In RDN, DGs efficiently minimize actual power losses and increase the voltage profile [27]. The ARDN concept is fundamentally a variant of RDN reconfiguration, allowing for more precise DG deployment and loss reduction. The inclusion of normally open switches (NOS) ensures the radial operation of DN [28]. To transfer loads across a community of interconnected radial feeders while maintaining synchronized power flow, the reconfiguration is accomplished by adjusting the open/closed status of tie/sectionalizing switches [29]. An abrupt increase in demand and capability deficiency of T&D facilities are the prime concerns, which ensure a rapid growth in deployment of DGs at distribution level [30]. As a result, ARDN encourages more precise deployment of DG units and offers many benefits, such as loss reduction and enhanced voltage profile. In respect of voltage sags and electricity supply problems, this precise DG placement feature improve the system’s stability and power quality issues. The schematic diagram for ARDN is shown in Figure 7.

3.2. Ring or Loop Distribution Networks

The loop distribution network (LDN) is a more modern variant of the traditional distribution network (DN) under the umbrella of advanced power distribution mechanism (APDS) applications in which main feeders operate in a loop structure [31]. In LDN topology the utility can deliver power in any direction of the loop [32]. Higher reliability, improved stability index, fault isolation and large DG penetration features make the LDN topology more preferrable over conventional RDN topology [31]. The schematic view of LDN is shown in Figure 8. The advantages of looping the traditional RDN by a series of electronic devices to monitor power flow were explored by the authors in [33]. In [34], the authors proposed a load transfer model (LTM) to investigate the effectiveness of transition to closed loop distribution network (CLDN). However, as opposed to conventional planning approaches, LDNs typically have higher informing cost limits, enhanced protection needs, sophisticated automation, anticipated increases in short-circuit currents, dynamic monitoring needs, and comparatively high losses. Costs and losses are two main constraints that restrict further research and development on LDN and encourage the adoption of RDN configuration at operation level. However, some entities, such as the Singapore Power Company, Taiwan Power Company, Hong Kong Electrical Supply Company, and others, continue to provide high-reliability service to their customers using LDNs. Need-based loops and ordinarily closed loops are two different operating topologies in LDN. Close a select number of tie switches to convert a component or a set of tie switches to convert the whole mechanism into a loop to accomplish either mode.

3.3. Meshed Distribution Networks

The meshed distribution network (MDN) is considered a variant of LDN, generally referred as a multiloop distribution network (MLDN). It is normally exposed to the same problems and, for the most part, the same specifications as LDN. The schematic illustration of MDN is shown in Figure 9. While a meshed network can be an economically suitable alternative, it does necessitate modifying the protection approaches, which is incompatible with a meshed operation. In comparison to LDN, it has the benefits of higher stability, large DG infiltration, better voltage response, and relatively lower power losses [34]. Comparably with RDNs, primary MDNs offer benefits such as lower short circuit current, shorter overall conductor length, and increased reliability and fault tolerance [36]. However, the practical implementation of MDN is not common globally due to its complex installation and coordination of relays during protection. Standard planning techniques, on the other hand, are more vulnerable to failures and early recovery due to dynamic fault traceability [37]. For standard protection purposes, based on fault current limiter devices, the NOS could be replaced by “quick deloopers” that can bring the system back to radial configuration for stable operation in cases of abnormal conditions [28]. In MDN, the most common configurations are weakly meshed and completely meshed DN [38,39].

3.4. Microgrids

Microgrids (MG), a subset of the SG concept and regarded as the building blocks of it, are possibly the most promising and innovative network structure for ensuring a reliable, cost-effective, and environment-friendly energy solution for a community. MGs allow and encourage a large penetration of DGs, such as microturbines, fuel cells, and PV arrays, as well as ESS, such as flywheels, supercapacitors, and adjustable loads, at the distribution level, which provide control capabilities over network operation. Figure 10 depicts the fundamental structure of MGs that contain both AC and DC topology. The main advantages and vital role of MGs is their operation in both grid-connected and stand-alone mode. These control skills enable DNs, which are often tied to the upstream DN, to operate when disconnected from the main grid in the event of faults or other external disruptions or catastrophes, therefore improving supply quality [7].

According to a European research project [7], “MGs are LV distribution systems that incorporate DERs (microturbines, fuel cells, PV, and so on) as well as storage devices (flywheels, energy capacitors, and batteries) and flexible loads. If linked to the grid, such systems can function in a non-autonomous mode, or in an autonomous mode if detached from the main grid. When controlled and coordinated properly, the functioning of micro-sources in the network can bring significant improvements to overall system performance.” According to [40], “MG concept implies a cluster of loads and micro-sources that operate as a single controlled system that delivers both electricity and heat to its immediate surroundings.” According to the US Department of Energy (DOE) [41], “the term ‘MG’ refers to a group of associated loads and DERs with well-defined electrical boundaries that behave as a single controlled entity with regard to grid and may connect and detach from utility grid to operate in both grid-connected and stand-alone modes.”

From distribution perspectives, MGs are classified into three types i.e., AC, DC, and hybrid [42]. In AC-MGs, all RESs and loads are linked to a shared AC bus, while DC-MGs can be operated in grid-connected or stand-alone mode. All the DC loads are connected directly with a DC bus. Where the hybrid MG is connected to utility grid through a static transfer switch (STS) or point of common coupling (PCC). The power electronic interface controls the power flow between the networks and the utility grid. The power direction is determined by the balance of load and generation.

3.5. Multi-Microgrids

According to IEEE Std 1547.4, a large DN can be modified to clustered MGs called MMGs to facilitate the control and operation (C&O) infrastructure in future DN [43]. With the MGs’ large-scale interconnection to the electricity grid, an MMG architecture will emerge from a group of nearby MGs in each region. Figure 11 depicts MMG architecture, which provides new approaches to deal with the challenges of planning and facilitate the control and operation infrastructure in future distribution scenarios while achieving certain goals, such as power quality, reliability, and DG siting and scaling.

MMGs offer an efficient P2P energy trading and information sharing facility between different MGs to achieve cost-reliable and efficient solutions of electricity delivery to end customers. This improves the economical operation for all stakeholders. When the permeability of MGs improves, it becomes more difficult to coordinate the efficient and dependable functioning of various sub-MGs. MMGs are classified into three constitutional frameworks: AC, DC, and hybrid. All the microsources in AC-MMGs are connected to AC buses via converters. Microsources are connected to DC buses in DC-MMGs via converters, which are typically loaded with DC loads, and the DC-MMGs are connected to the external power grid via inverters [44]. Remote places with limited access to utility grid electricity, family communities, business buildings, industrial parks, and experimental research are some of the main application possibilities for MMGs.

Figure 11. Multimicrogrid: a conceptual and graphical overview (adapted from [45]).

Figure 11. Multimicrogrid: a conceptual and graphical overview (adapted from [45]).

3.6. Virtual Power Plants

Virtual power plants (VPPs) have emerged as a potential option for flexible energy management in ADNs, with the rising penetration of DERs such as RESs, battery storage, and controlled loads [46]. Different authors treat the definition of virtual power plants (VPPs) in different ways in the literature. For example, [47] defined the VPP as an autonomous MG. A VPP, according to the FENIX project concept, is the aggregation of many heterogeneous DERs with varying capacities, which generates a single operational profile from a composite of the attributes describing each DERs and can integrate the network’s influence on aggregate DER output for the purpose of enhancing reliability, power generation, and trading or selling of power on the electricity market [48,49]. Fundamentally, the VPP concept is comparable to the MMG approach, which is controlled and managed by a central body and consists of a variety of DERs with responsive loads dispersed throughout vast geographic areas [12]. The main difference between MGs and VPPs is that MGs can be either grid-connected or stand-alone, while VPPs will always operate in grid-connected mode. On the basis of operational and technical constraints, VPPs are divided into two major classes in the literature, i.e., commercial (focusing solely on the economic aspect, with no requirement for geographic closeness of aggregated sources) and technical (in contrast to the commercial VPP, aggregated resources are physically close together, and their coordinated control takes into account electrical network constraints) [50]. Figure 12 shows the conceptual framework for VPP.

3.7. Smart Homes

Smart home (SH) service is an essential component of smart grid usage. It is a real-time interactive response between the power grid and users that improves the power grid’s comprehensive service capability while also realizing intelligent and interactive electricity use. It also improves the power grid’s operation mode and users’ consumption patterns to improve end users’ energy efficiency [52]. The SH concept can be defined from many different perspectives. Different authors define SHs differently, but the conceptual identification of SHs from all studies’ point of view remains the same. According to [53], the concept of SHs is defined as “SH is a pervasive computing application in which the home environment is supervised by ambient intelligence in order to deliver context-aware services and enable remote home management.” SHs, according to [54], are residences equipped with computing and information technology that foresees and responds to the needs of the residents, employed to indorse their ease, expediency, security, and entertainment through the management of technology within the home and influences on the outside world. According to another definition, “SH is a residential-based platform that connects various facilities through the network to meet the system’s automation requirements and provide more convenient control and management. It makes use of internet of things (IoT), computer, control, image display, and communication technology [52].” As a concept, SHs have significant appeal in the present and future SG domains due to their ability to optimize power usage based on electricity consumption and production capacity. Internal network, intelligent control, and home automation are the key enablers of SHs. Services offered by SHs include comfort, remote access, home automation, repository, energy optimization, and health care. SHs demand specialties from different background, e.g., communications engineers, computer and software engineers, and thermal engineering specialists. However, power and energy specialists are vital in designing and optimizing the framework of SHs, so a lot of research has been carried out in the last few years in the field of power and energy.

From a power and energy specialists’ viewpoint, the research carried out in the last few years in planning, operation, and control aspects of SHs is discussed below. The Figure 13 shows a conceptual framework for SHs.

3.8. Smart Neighborhoods

Buildings may be a huge source of flexibility when it comes to dealing with the SG’s problems. A cluster of buildings or a neighborhood, in particular, can provide more flexibility by adjusting their energy profile. The interactions between buildings and the SG in a neighborhood have the potential to assist in controlling energy consumption in the SG environment. These interactions between SG and SNH may be achieved using the MAS paradigm, with the SNH acting as an MG. SBs and their related controls in a neighborhood are capable of reacting to DR requests from system operators to manage peak demand in order to minimize demand charges or to alter operations depending on real-time market prices of electricity. At whatever level, the operation of a SNH has two primary goals: to maximize energy flexibility and to maximize user comfort [56]. SNHs [57] are high-performance homes constructed with increased energy-efficiency features that exceed industry requirements. Features of SNH homes include energy efficiency, interaction between appliances, novel security and privacy solutions and home automation. The concepts of SNHs is shown in Figure 14.

Standard features of SNH include programmable thermostats, better insulation, high-efficiency heat pumps, and water heaters and appliances. SNH homes are designed based on the homeowner’s demands and mindset. They are outfitted with smart technologies and give a suitable network to assist in managing and controlling various parts of the home.

3.9. Smart Buildings

The concept of SBs originated in the early 1980s. In a 1984 New York Times story, real estate developers were characterized as constructing “a new generation of buildings that virtually think for their own, termed intelligent buildings.” Such types of intelligent buildings were a combination of old structure building management and telecommunication technologies [58]. SBs are a structure that automatically controls the different operations of buildings, including HVAC systems, lightning, and security. The topic of SBs is inextricably related to the smart grid concept. SBs rely on a collection of technologies to improve energy efficiency, user comfort, and facility monitoring and security. SBs by definition can combine building management needs with IT technologies to enhance system efficiency, at the same time simplifying typical service functions. IoT-based technologies are the key drivers used in building management systems (BMSs) and monitor HVAC systems snd software packages for automatically switching off/on the different appliances [58].

In the context of the smart grid concept, the following features are the key enablers of SBs [59], shown in Figure 15.

Renewable Energy Resources: SBs consist of RESs (mostly PV systems) installed on the roof, parking, or ground of the building along with BES system. These RESs operate in such a way that during peak time slots when energy prices are high, a building operates only on solar-based system. During low demand and low energy price hours, the building takes electricity from the utility to meet its demand, while the PV system charges the battery bank for upcoming peak hours. RESs may be coordinated via smart meters to offer optimal energy flow to balance out the system’s frequency level. Net metering allows energy to be sold to the grid or utilized by the building.

Energy Storage: Electric storage systems provide backup during peak demand. Battery backup gets charged during off-peak time slots through RESs and discharged during peak time slots. Battery backup systems could be either BESs or EVs (V2G concept). EVs can act as ESS. EVs charge during off-peak time slots (G2V) and discharged and provide the energy for internal consumption of building (V2G).

Smart Meter: Smart meters allow for bidirectional communication and remote reading. The building has real-time consumption and energy pricing information depending on each interval of time. Advanced metering infrastructure (AMI) enhances automatic meter reading (AMR) technology by enabling bidirectional communication allowing instructions for a variety of applications such as time-based pricing information and DR actions.

Smart Appliances: Advanced communication technologies offer interoperability of the smart appliances in the building, where devices can communicate with each other for real-time decision-making purposes. Smart appliances can be used as part of building automation or on their own. They can monitor the building’s conditions and switch off or on based on user-defined conditions.

Building Automation: Building automation consists of sensors, actuators, controllers, central unit, interface, and a network standard for communication. It allows users to program the behavior of a building depending on predefined circumstances.

Broadband Connection: Broadband connectivity allows communication with utility and other SBs via broadband over power lines (BPL), Worldwide Interoperability for Microwave Access (WiMAX), Global System for Mobile Communications (GSM), or other communication standards, allowing the development of active MG or VPP under the smart grid paradigm.

3.10. Smart Cities

During the last two decades, the term “smart cities” has received specific attention and grown more common in scientific literature and global policies [60]. From the definition perspective of SCs, nor single framework is adopted. Many definitions have been reported in the literature. In the 1990s, the phrase “SCs” was introduced for the first time. At the time, the emphasis was on the importance of new ICT in terms of contemporary city infrastructures. By substituting other adjectives for “smart,” such as “intelligent” or “digital,” a variety of conceptual variations might be created. Detailed definitions of SCs carried in different research studies are presented in [60]. According to [61], an SC is one in which ICT is combined with conventional infrastructures and is coordinated and integrated utilizing modern digital technology. San Diego, San Francisco, Ottawa, Brisbane, Amsterdam, Kyoto, and Bangalore, among other SC pioneers, are now setting the standard for others to follow [62]. SCs facilitate political efficiency and social and cultural development according to their characteristics. SCs place a strong focus on business-led urban development and innovative initiatives to help cities thrive [60]. The SC idea provides possibilities to address various challenges due to rapid urbanization, alleviate urban challenges and provide a better living environment for residents [21,63], but the concept brings several challenges, and is still under study by different research groups across the globe. Cities consume a large share of the overall utilization of energy, accounting for more than 75% of global energy production and 80% of GHG emissions [64].

Various research has been conducted in this subject in order to arrive at an SC model with an average technological size, which is interconnected and sustainable, as well as comfortable, appealing, and secure for the community. SCs are an important application of IOT. IOT helps in construction of SCs by application handling, data management and processing, and data generation and acquisition [65]. According to Cities in Motion Index (CIMI) results, the top 10 SCs are New York (USA), London (UK), Paris (France), San Francisco (USA), Boston (USA), Amsterdam (Netherlands), Chicago (USA), Seoul (South Korea), Geneva (Switzerland), and Sydney (Australia). The concept of SCs is shown in Figure 16.

SCs and SGs: Relationship: SCs are a rational expansion of the SG idea, and their implementation is inextricably linked to the modernization of existing power systems [68] SGs modernize power systems by including self-healing, automation, and monitoring, and keeps end users aware of their energy consumption and encourage them to modify their energy consumption patterns in response to electricity prices. The SG is the nerve center of the SC, which would not be complete without it. SCs rely on an SG to ensure reliable energy transfer for their many operations, as well as to provide possibilities for sustainability, efficiency improvements, and, most pertinently, to allow cooperation between different stakeholders involved in this concept [69]. World’s top 10 SCs are shown in Figure 17.

4. Active Radial Distribution Network

4.1. From Planning Perspectives

During recent years, optimal planning of an active distribution network gained wider attention. Due to their simple structure and topology, radial distribution networks are adopted widely across the globe. From basic concepts, ARDN is a reconfigured variant of RDN that allows the participation of DGs and loss minimization at the user end. However, when DGs are introduced, this poses some technical and operational challenges on the system. For that purpose, optimal planning of RDN reconfiguration is of prime importance for the reliability and stability of the system within acceptable bounds. Optimal planning for RDN reconfiguration in terms of power balance, resource allocation (DGs and compensation device types, size, and location), and topology reconfiguration is discussed below based on the research. The reviewed work of ARDN from planning perspectives is summarized in

Table 1. Active radial distribution networks: reviewed literature from planning perspectives.

Ref.	Considered DN	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Aspects Covered
[70,71]	ARDN	Generation and load forecasting	1. (⇊) Capital investment 2. (⇊) Operation cost 3. (⇊) Levelized energy cost 4. (⇈) Safety	1. Geog. constraints 2. Elect. constraints 3. DGs capacity 4. DGs loading 5. Seasonal load 6. P and Q limits 7. Customer info.	1. Simple DN 2. MV-DN in Portugal	1. Long-term planning 2. Short-term planning		1. Wind 2. Fuel cell 3. PV 4. Grid --	Techno-economic
[72]	ARDN	Generation and load forecasting	3. (⇊) Weighted squared residuals 3. (⇊) Forecasting error	1. and P&Q 2. Load 3. State estimation	18-nodes test network	Long-term planning			Technical
[73,74]	ARDN	Generation and load forecasting	1. Load prediction 2. Uncertainty observation 3. (⇈) Voltage profile 4. (⇊) Power mismatch	1. Capacitor sizing 2. Load forecasting 3. P and Q limits 4. Load data	33/11 kV substation near Kakatiya university 34 and 69-nodes RDN	Short-term planning			Technical
[75,76,77,78,79]	ARDN	Generation and load forecasting	1. (⇈) Voltage profile 2. (⇈) Safety 3. (⇊) Prediction error 4. (⇈) Renewable penetration 5. (⇊) Power loss	1. Rated power 2. No. of customers 3. Client types 4. Load data 5. Load forecasting 6. Energy consume 7. Network reconf.	OSIRIS pilot project with 24-MV feeders Bonneville Power Administration (BPA) balancing area IEEE 123-bus test system	Short-term planning		-- -- -- Wind --	Technical
[80,81,82,83,84]	ARDN	Resource allocation DGs (location + size)	1. (⇈) Voltage profile 3. (⇊) Voltage deviation 2. (⇊) Real PL 3. (⇈) DG penetration	1. Penetration level 2. Islanding 3. Voltage limits 4. Voltage stability 5. Power flow limit 6. P and Q limits	1. IEEE 33 and 69 RDN 2. 33 and 69-bus DN 3. IEEE 33, 66-bus DN 4. 12, 33, 69-bus RDN 5. IEEE 15, 33-bus DS	Optimization based operational planning			Technical
[85,86]	ARDN	Resource allocation DG (placement + sizing)	1. (⇊) Power loss 2. (⇈) Voltage profile 3. (⇊) Voltage deviation	1. Penalty factor 2. Feeder constraints 3. V&I limits 4. Power flow limit 5. Power balance 6. DG power limits 7. Varying loading	IEEE 33-bus RDS IEEE 33-bus, 69-bus, and 118-bus RDS	Multiobjective planning			Technical
[87,88,89,90]	ARDN	Resource allocation DG (size + location)	1. (⇊) Energy loss 1. (⇊) Active PL 2. (⇈) Voltage profile 2. (⇊) Energy loss 2. (⇊) Loss expenses	1. Power balance 2. Voltage limits 3. Line current limits 4. P&Q limits 5. VSI 6. Thermal limits	1. 33-bus, 69-bus, and 118-bus RDN 2. IEEE 33-bus and 66-bus DS 3. IEEE 33-bus, 69-bus, and 118-bus RDS	Long-term (1 year) planning		Biomass Wind Solar PV
[91,92]	ARDN, MDN	Resource allocation SCs (siting + sizing)	1. (⇊) Total active PL 2. (⇊) Energy cost 3. (⇈) Voltage profile 4. (⇊) PL 5. (⇊) Operating cost 6. (⇊) Investment capacitor costs	1. Power flow cons. 2. V and Q limits 3. Capacitor’s constraints 4. Load flow const.	1. 69, 85-bus and Real DS 2. IEEE 34, 85-bus RDN 3. 10, 34, and 85-bus and CIVANLAR MDS	Multiobjective (1-year) operational planning	1. COA 2. SSA and LS WOA 3. MILP
[93]	ARDN	Resource allocation Conductor (type + sizing)	1. (⇊) Conductor capital cost 2. (⇊) Conductor energy loss cost	1. Bus voltage limits 2. Conductor current capacity 3. Power flow cons.	16-bus and 85-bus test system	Yearly operational planning	CSA
[94]	ARDN	Topology selection	1. (⇊) Active power loss 2. (⇈) Reliability	1. Radiality condition 2. and P&Q flow 3. No of customers	33-nodes DS of Taiwan	Multiobjective planning	e-constraint method using lexico-graphic optimization
[95,96,97]	ARDN	Topology selection	1. (⇊) Power loss 2. (⇈) Voltage profile	1. and P&Q limits 2. Branch current limit 3. Branch voltage limit 4. Voltage deviation 5. Power flow 6. Radiality	1. 33-bus IEEE-RDS 2. IEEE 16, 33 and 69-bus test system 3. 70-node DS	Multiobjective planning	1. BIBC-BCBV method and BPSO 2. GA 3. Fuzzy multiobjective method 4. MPSO
[98,99,100]	ARDN	Topology selection	1. (⇊) Power loss 2. (⇈) Relibility	7. Node voltage limit 8. Feeder cap. Limit 9. Radiality 10. and P&Q limits 11. and V&I limits 12. DG power limits 14. Kirchhoff’s and V&I	1. 33 and 69-bus test DS 2. IEEE 33 and 89-nodes DS 3. 69 and 136-bus system	Single-objective planning	1. MPSO 2. Hyper-spherical search (HSS) algorithm 3. GA

.

4.1.1. Generation and Load Forecasting

The planning of an optimum distribution system is a difficult task, with several objectives and limitations. With uncertainties in both load and generation, the situation becomes more complicated in smart distribution grids, so uncertainty is the fundamental and basic feature of power system planning challenges. This problem is exacerbated in smart distribution grids by errors in demand and generation forecasts. Various research groups applied different techniques to investigate the uncertainties related to load and generation radial configuration-based distribution power systems. The authors of [70] proposed an optimal smart distribution grid multistage expansion planning model based on point estimation, taking into account reinforcement or installation time, capacity, and placement of MV substations and DERs. Based on neural networks, [101] presented an extensive comparative model of forecasting methods in a radially configured power system. Different approaches, such as combined regression models and artificial neural networks (ANN) [71,72,73,74], adoptive load forecasting (ALF) [75], multilayer bidirectional recurrent neural network (BRNN) [76], deep learning (DL) and image encoding (IE) [77], and gray system theory (GST) [78], have been reported in the literature for load forecasting to enhance the reliability and efficacy of radial configuration-based distribution networks. An extensive and comprehensive review was carried out [102,103] to investigate the different load forecasting models and techniques for distribution network. A novel distribution system network reconfiguration based short-term load forecasting technique was developed in [79] for an optimal topology design at every scheduled time step.

4.1.2. Resource Allocation

When addressing ARDN network planning issues, planning engineers need to deal with feeder consolidation, optimal DG placement, optimal placement of reactive power compensation (shunt capacitors, SATATCOM and FACT, etc.) devices, and network reconfiguration concerns. Choosing appropriate DG characteristics and DG parameters is a major problem for distribution system planners to get the most out of the DG unit [30]. In the planning stage, the optimum positioning and sizing of DG in the ARDN is critical. The proper siting of DG will result in a number of advantages, including reduced active power losses, increased network efficiency, and enhanced voltage profile [104]. The system’s continuity in islanding mode of operation can results in various technical problems. Thus, based on a two-stage genetic algorithm (TSGA), an extensive study has been conducted to reduce the consequence of islanding while improve the entire voltage profile and reducing the active power (P) losses through optimal DG scaling in RDN [80]. Researchers in [81,82,83] developed a novel and simple computational via basic mathematical model to optimize DN power loss by planning optimal capacities, locations, and power factor for each DG unit in the test system. Different techniques, such as multiobjective particle swarm optimization (MOPSO) [85], hybrid sequential quadratic programming (HSQP) [86], krill herd and opposition-based learning (KH&OBL) [87,88], intersect mutation differential evolution [89], teaching learning-based optimization (TLBO) [90], and grey wolf optimization (GWO) [105], have been reported in the literature for minimizing different types of losses in the system and maximizing the reliability of the system through optimal placement and sizing of DGs.

As reactive compensation is directly proportional relation with voltage compensation, so a better compensation of Q can enhance the voltage profile of system. In terms of P&Q compensation, DGs are classified as [106]: category 1—DGs that are capable of supplying only P (e.g., PV unit), category 2—DGs that are capable of supplying only Q (e.g., Syn.Comp), category 3—DGs that can supply P, but absorb Q (e.g., Ind.Gen), and category 4—DGs that can supply both P&Q to the system (e.g., Syn.Gen). The authors in [84] considered category 4 DGs and proposed a hybrid analytical and PSO (HA&PSO) approach for voltage profile improvement and at the same time optimizing the DG injection level and distribution line losses in the system. In [91,107], a hybrid optimization approach (HOA) based on salp swarm algorithm (SSA) plus loss sensitivity and whale optimization algorithm (WOA) indices were proposed to evaluate the optimum distribution and sizing of shunt capacitors in RDN, respectively. With the aim of minimizing network power losses, an optimal placement of DSTATCOM using the harmony search algorithm (HSA) has been carried out in [108]. Different approaches, such as Hornoy search algorithm (HSA), improved HAS, sensitivity analysis, memetic algorithm, ETAP, particle artificial bee colony algorithm (PABC), differential evolution algorithm, cuckoo search optimization algorithm, mine blast algorithm, and hybrid particle swarm techniques, have been reported in the literature for optimal placement and sizing different compensation devices in the radial distribution network [109,110,111,112,113,114,115,116,117,118,119,120,121]. Finally, [122] provides a comprehensive overview of the statistical approaches used for the optimum collection and placement of reactive power compensating components in distribution systems. Optimal selection of conductor size is a very relevant concern of planning engineers. A novel method has been proposed based on the crow search algorithm (CSA) for distribution of optimal conductor size selection in a radial distribution network [93].

4.1.3. Topology Selection

Reconfiguration of a radial distribution network allows one to modify the topology of the framework to enhance the stability and reliability indices during high penetration of RESs. In [94], optimal radial design was achieved by reducing active power losses and a set of frequently used reliability indices in the context of multiobjective ARDN planning. Similarly, binary particle swarm optimization (BPSO) [95], improved genetic algorithm (IGA) [96], fuzzy multiobjective method [97], modified particle swarm optimization (MPSO) [98], and soft open points (SOP) method [99] have been carried out by different research groups for optimal reconfiguration of radial distribution systems. A distribution network reconfiguration approach based on IGA is proposed for both indices of power loss reduction and reliability enhancement [100]. Based on an improved selective BPSO approach, an extensive study was conducted in [123] for optimal reconfiguration of radial distribution network from the perspective of power loss reduction.

5. Ring or Loop Distribution Network

5.1. From Planning Perspectives

LDN allows greater reliability over radial topology under the application of advanced power distribution systems. However, to achieve sufficient reliability requires high cost and other planning concerns due to higher upgrade and protection constraints in cases of large REG penetration. For that purpose, optimal planning of LDN has a significant effect for achieving a better reconfigured network for future smart distribution networks. The reviewed literature for LDN from planning perspectives is summarized in

Table 2. Loop distribution networks: reviewed literature from planning perspectives.

Ref.	Considered DN	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Aspects Covered
[124]	LDN	Fault location	1. (⇈) Precise fault location performance	1. Sequence components 2. Feeder capacities 3. Capacitances 4. Inductances 5. Transformer’s topologies	10 kV, 16-node test feeder	Operational planning		--	Technical
[125]	LDN	Network reconfiguration	1. (⇈) Voltage profile 2. (⇊) Power loss	1. Power flow balance 2. Bus voltage limits 3. Line current limits 4. RN structure	33-bus, 84-bus, 119-bus, and 136-bus DS	Operational planning		--	Technical
[126]	RDN and LDN	Network reconfiguration	1. (⇈) Voltage profile 2. (⇈) Loadability	1. Power flow balance 2. Bus voltage limits 3. Line current limits 4. and P&Q limits 5. Power capacity limit	IEEE 33- and 69-bus test systems	Operational planning		--	Technical
[127,128]	LDN	DG allocation (sizing and siting)	1. (⇈) Voltage profile 2. (⇊) Power loss	1. Power flow limit 2. Voltage limit 3. DG capacity limit 4. T/F rating limit 5. Power flow eqs. 6. Current limits	1. 15-node test feeder of Egypt 2. Riyadh city MV network	1. Long-term (5 years) planning 2. Operational planning		1. Hydel 2. Wind 3. Solar PV	Technical
[129]	LDN	DG allocation (siting)	1. (⇈) Voltage profile 2. (⇊) Power loss 1. (⇈) Loadability 1. (⇈) Reliability	1. Voltage stability index 2. Voltage limits 3. Power angles 4. P and Q limits 5. R/X ratios	Korean Electric Power Corporation DN and IEEE 69 bus test DN	Operational planning		--	Technical

.

5.1.1. Generation and Load Forecasting

For a neutral ineffectively grounded loop distribution scheme, [124] presents the distribution of loop closing zero sequence current using the single-phase grounding (SPG) fault location technique. With the intention of assessing efficiency for a normally closed-loop distribution network, [130] applied the pilot relaying approach to a protection scheme to reduce the number and the length of fault-induced interruptions.

A Chaotic Stochastic Fractal Search Algorithm (CSFSA) was implemented in [125] with the aim of minimizing power loss and improving voltage efficiency while simultaneously solving the reconfiguration problem in DN with three common topologies: radial, loop, and mesh. With proper assistance of continuation power flow (CPF) approach, the load-ability of the LDN was evaluated against voltage stability limits in [126].

5.1.2. Resource Allocation

In [131], the authors proposed a new planning framework for DG placement and sizing in LDN focused on the voltage stability index (VSI) and enhanced loss minimization (LM) formulations. After measuring the fault position in LDN with DG, [132] suggested a protection scheme to isolate the faulty section from the main supply. In [127], a switching technique is presented to achieve the optimum siting and sizing of DG during the restoring fault cycle in an LDN. The authors of [133] developed a customized continuous power flow approach based on loop-analysis-based power flow (LBPF) analysis to analyze voltage stability for both RDNs and MDNs. The authors of [128] mixed two approaches to find the optimal PV-based DG allocation and sizing for an LD underground feeder. The authors in [134] used multicriteria decision analysis (MCDA) to determine the best location for DG in order to maximize various parameters of attention. In [129], the authors proposed a new voltage stability index (VSI) for normally closed LDN (NCLDN) to signify voltage sensitive nodes. The authors in [135] suggested using bidirectional overcurrent relays and GOOSE communications to increase the security of MV closed loop-configured DN.

6. Meshed Distribution Networks

6.1. From Planning Perspectives

The impact of DER connectivity is not taken into account when traditional electric power distribution networks are constructed. This frequently leads to operational circumstances that would not exist in a traditional system. Reference [136] addressed a number of system concerns that may arise when DERs make a large penetration into DNs. A transformation of the radial topology of primary feeders to meshed may greatly enhance the facility to penetrate more DERs into the network. GA-based multiple-objective indicators were presented in [137] for improving primary feeders from a normally closed arrangement to a meshed arrangement for DG installation. The impacts of different load models by butterfly particle swarm optimization (BF-PSO) on optimal planning of DGs in DN were examined by [138]. Different techniques, such as unique multistage expansion planning [139], imperialist competitive algorithm (ICA) [140], and heuristic algorithm (HA) [141,142], have been reported in the literature for optimal network expansion planning of MDN under different constraints.

6.1.1. Generation and Load Forecasting

The optimization method’s core content and process are comparable to existing MDN planning. It consists of collection of data, power supply mesh and unit division, power grid condition analysis, load forecasting, target grid planning and transitional scheme determination, and distribution automation planning, etc. [143]. The estimation of different load models’ parameters helps in load forecasting under future scenarios [144]. Different state estimation techniques have been investigated in the literature from the perspectives of load forecasting under different scenarios. The authors of [145] proposed a load forecasting-based real-time state estimation (SE) technique for radial and meshed DNs that use consecutive weighted least squares estimate and power flow to modify forecasted load values to match real-time measurements. Based on spatial load forecasting and the natural growth method, [143] carried out long-term, short-term, and medium-term forecasting for the division of power supply in MDN. The flow of optimum distribution system is shown in Figure 18. The reviewed literature for MDN from planning perspectives is summarized in Table 3.

Table 3. Meshed distribution networks: reviewed literature from planning perspectives.

Ref.	Considered DN	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Aspects Covered
[137,146]	MDN	Generation and load forecasting	1. (⇊) Voltage deviation 2. (⇊) Power loss 3. (⇊) Line losses 4. (⇈) DG capacity	1. P and Q limits 2. Bus voltages limits 3. DG capacity limits 4. Power flow limits	1. IEEE 30-bus 2. 28-bus test feeder	Operational and multiobjective planning		1. Synch. DG 2. Wind power generator	Technical
[143,147]	MDN and RDN	Generation and load forecasting	1. (⇈) Reliability 2. (⇊) Voltage deviation 3. (⇊) P and Q loss	1. P and Q limits 2. Bus voltages limits 3. Power flow limits 4. Voltage angle limits 5. DS capacity limits 6. Power factor limits	1. MV-PG of Hubei province, China 2. IEEE 33-bus and IEEE 69-bus RDS	1. Operational and multiobjective planning 2. Long-term (10 years) planning		--	Technical
[145,148]	MDN	Generation and load forecasting	1. (⇈) Load scale 2. (⇊) Voltage deviation 3. (⇊) Power loss	1. P and Q limits 2. Bus voltages limits 3. Bus V-angle limits 4. Power flow limits	1. Two IEEE-123 Sub-Networks 2. IEEE 33 and 69-bus system	Operational and multiobjective planning		--	Technical
[92,149,150,151,152]	RDN and MDN	Resource allocation (capacitor sizing and siting)	1. (⇈) Vol. profile 2. (⇊) Power loss 3. (⇈) Penetration 4. (⇊) Cost	1. P and Q limits 2. Load growth 3. Voltage stability 4. Bus voltages 5. Power capacity	1. IEEE 33-bus sys. 2. 69-bus MDS 3. 10, 34, and 85-bus RDS, CIVANLAR 4. UK meshed DS 5. 33 and 69-bus MS	Long-term planning		1. DG 2. D-STATCOM 3. Shunt Capacitor 4. D-STATCOM 5. DGs and D-STATCOM	Techno-economic
[39,153]	MDN	Resource allocation (DG’s size and siting)	1. (⇈) System loading margin 2. (⇈) DISCOs Profit	1. Bus voltage 2. Feeder PF	1. 6-bus and 30-bus system 2. 9-bus, 14-bus, and 30-bus system	MO planning		-- Generator and synch. condenser	Techno-economic
[154,155,156,157]	MDN	Resource allocation (DG’s size and siting)	1. (⇈) Volt. Profile 2. (⇊) Loss 3. (⇊) Cost	1. Power factor 2. Load flow 3. Voltage limits 4. DG penetration	1. 38-bus UK DS 2. IEEE 69-bus sys. 3. 33-bus RDN and CIVANLAR-MDS	Short-term (1-year) planning Operational planning		--	Techno-economic
[158,159]	MDN	Topology selection	1. (⇊) Loss 2. (⇈) Penetration 3. (⇈) Voltage regulation 2. (⇈) Resiliency	1. X/R ratio and P cap. 2. Bus voltages limits 3. Thermal loading 4. Line currents limits 5. P and Q limits	1. 11 kV distribution network 2. 37-bus standard test bench	1. Short-term (yearly) planning 2. Operational planning		1. DG (DFIG Wind generator) 2. PV, WT, and Shunt capacitor	Technical
[160,161]	MDN	Topology selection	1. (⇈) Load supply 2. (⇊) Number of witching operations 3. (⇊) Loss	1. Load bal. and thermal 2. PF and phase angle 3. V, P, and Q limits 4. RN topology	1. 32, 70, 135, and 880-buses network 2. 5, 33, and 70-nodes test system	Operational planning		--	Technical
[162]	MDN	Topology selection	1. (⇊) Total cost (investment cost, reliability cost, and maintenance cost)	1. Distflow equations 2. Power flow 3. Node voltages 4. Capacity available 5. Switching and feeder	54-node system	Long-term (10 years) planning		--	Economical
[163]	MDN	Topology selection	1. (⇊) Loss 2. (⇊) Vol. deviation 3. (⇊) DG gen. loss	1. P and Q limits 2. Voltage limits	IEEE 33-bus DS	MO planning		PV	Technical

Table 3. Meshed distribution networks: reviewed literature from planning perspectives.

Ref.	Considered DN	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Aspects Covered
[137,146]	MDN	Generation and load forecasting	1. (⇊) Voltage deviation 2. (⇊) Power loss 3. (⇊) Line losses 4. (⇈) DG capacity	1. P and Q limits 2. Bus voltages limits 3. DG capacity limits 4. Power flow limits	1. IEEE 30-bus 2. 28-bus test feeder	Operational and multiobjective planning		1. Synch. DG 2. Wind power generator	Technical
[143,147]	MDN and RDN	Generation and load forecasting	1. (⇈) Reliability 2. (⇊) Voltage deviation 3. (⇊) P and Q loss	1. P and Q limits 2. Bus voltages limits 3. Power flow limits 4. Voltage angle limits 5. DS capacity limits 6. Power factor limits	1. MV-PG of Hubei province, China 2. IEEE 33-bus and IEEE 69-bus RDS	1. Operational and multiobjective planning 2. Long-term (10 years) planning		--	Technical
[145,148]	MDN	Generation and load forecasting	1. (⇈) Load scale 2. (⇊) Voltage deviation 3. (⇊) Power loss	1. P and Q limits 2. Bus voltages limits 3. Bus V-angle limits 4. Power flow limits	1. Two IEEE-123 Sub-Networks 2. IEEE 33 and 69-bus system	Operational and multiobjective planning		--	Technical
[92,149,150,151,152]	RDN and MDN	Resource allocation (capacitor sizing and siting)	1. (⇈) Vol. profile 2. (⇊) Power loss 3. (⇈) Penetration 4. (⇊) Cost	1. P and Q limits 2. Load growth 3. Voltage stability 4. Bus voltages 5. Power capacity	1. IEEE 33-bus sys. 2. 69-bus MDS 3. 10, 34, and 85-bus RDS, CIVANLAR 4. UK meshed DS 5. 33 and 69-bus MS	Long-term planning		1. DG 2. D-STATCOM 3. Shunt Capacitor 4. D-STATCOM 5. DGs and D-STATCOM	Techno-economic
[39,153]	MDN	Resource allocation (DG’s size and siting)	1. (⇈) System loading margin 2. (⇈) DISCOs Profit	1. Bus voltage 2. Feeder PF	1. 6-bus and 30-bus system 2. 9-bus, 14-bus, and 30-bus system	MO planning		-- Generator and synch. condenser	Techno-economic
[154,155,156,157]	MDN	Resource allocation (DG’s size and siting)	1. (⇈) Volt. Profile 2. (⇊) Loss 3. (⇊) Cost	1. Power factor 2. Load flow 3. Voltage limits 4. DG penetration	1. 38-bus UK DS 2. IEEE 69-bus sys. 3. 33-bus RDN and CIVANLAR-MDS	Short-term (1-year) planning Operational planning		--	Techno-economic
[158,159]	MDN	Topology selection	1. (⇊) Loss 2. (⇈) Penetration 3. (⇈) Voltage regulation 2. (⇈) Resiliency	1. X/R ratio and P cap. 2. Bus voltages limits 3. Thermal loading 4. Line currents limits 5. P and Q limits	1. 11 kV distribution network 2. 37-bus standard test bench	1. Short-term (yearly) planning 2. Operational planning		1. DG (DFIG Wind generator) 2. PV, WT, and Shunt capacitor	Technical
[160,161]	MDN	Topology selection	1. (⇈) Load supply 2. (⇊) Number of witching operations 3. (⇊) Loss	1. Load bal. and thermal 2. PF and phase angle 3. V, P, and Q limits 4. RN topology	1. 32, 70, 135, and 880-buses network 2. 5, 33, and 70-nodes test system	Operational planning		--	Technical
[162]	MDN	Topology selection	1. (⇊) Total cost (investment cost, reliability cost, and maintenance cost)	1. Distflow equations 2. Power flow 3. Node voltages 4. Capacity available 5. Switching and feeder	54-node system	Long-term (10 years) planning		--	Economical
[163]	MDN	Topology selection	1. (⇊) Loss 2. (⇊) Vol. deviation 3. (⇊) DG gen. loss	1. P and Q limits 2. Voltage limits	IEEE 33-bus DS	MO planning		PV	Technical

Figure 18. Flowchart for optimum distribution system (adapted from [164]).

Figure 18. Flowchart for optimum distribution system (adapted from [164]).

The authors in [149] proposed integrated decision-making planning (IDMP) from a multicriteria decision-making (MCDM) method for loss reduction by optimum asset planning in MDN to find a back-and-forth solution under various loading conditions for voltage stability evaluation.

6.1.2. Resource Allocation

For optimal allocation (sizing and siting) of different resources to enhance the reliability and efficacy of MDN with the objectives of power loss, annual cost minimization and voltage profile enhancement, different planning approaches, such as breeder and traditional genetic algorithm (BGA and GA) [39,146] have been reported in the literature. In [153], several techniques were examined to identify and address some rules for DG integrated mesh type networks. Some other approaches, such as multiobjective-based (MOGA) [137], sensitivity method [154], voltage stability index (VSI), MINLP [155,156], nonlinear programming (NLP), power loss sensitivity (PLS) [147], PSO [165], enhanced voltage stability assessment index (VSAI), loss minimalize condition (LMC) approach [148], multicriteria decision analysis (MCDA) [157], integrated decision-making planning approach including VSAI, loss minimization condition, and multi-criteria decision making (MCDM) [149], and variational algorithm [150], have been reported by different research groups with the aim of power loss and cost minimization, energy loss minimization, and voltage profile enhancement. By overcoming the limits of traditional Shapley value combinatorial explosion, [38] suggested a loss allocation (LA) approach to offer an empirical alternative for both RDN and MDN.

Similarly, from perspectives of compensation device planning, the authors in [92] proposed a mixed integer nonlinear programming (MINLP)-based approach to evaluate the optimal capacitor placement and sizing in a radial/mesh distribution network (R/MDN). The authors in [166] formulated the optimal placement of D-STATCOM in MDN using sensitivity approaches under probabilistic data of load in each season. The authors of [151] proposed a new enhanced voltage stability assessment index (VSAI-B)-centered planning method with synchronized optimal utilization of DG and D-STATCOM assets in an MDN, with the goal of voltage steadying, loss reduction, and related objectives.

6.1.3. Topology Selection

Meshed distribution networks (MDN) may have a favorable impact on power losses, voltage control, network reliability, and line and substation utilization, especially when large amounts of DG must be incorporated. It is of worth mentioning that with high DG penetration in DNs leave fewer reasons to continue to radial mode of operation and the higher the incentives to adopt meshed topology [167]. Network reconfiguration offers a more suitable topology/structure design to enhance the resilience of the distribution network. The authors in [158] arranged and enhanced the structure of existing radial topology to meshed network topology with the aim of reducing network losses and maximize the penetration for renewable-based DGs. For combined radial and meshed topologies, [160] presented a convex relaxation to solve reconfiguration problems, such as minimizing power loss, load harmonizing and power supply refurbishment in the network. Finally, [168] provides a brief overview of the effects of DG on RDN and MDN topologies at various voltage scales, with regard to various planning and operational problems. The authors in [159] proposed a three-stage self-healing algorithm (SHA) to restore maximum priority losses (PLs) through network reconfiguration for enhancing network resilience. Similarly, selective firefly algorithm (SFA), reliability constrained multistage expansion planning model (RCMEP), PSO, gossip-based algorithm, vector clock technique (VCT), and mesh adaptive direct search (MADS) have been reported in the literature [139,154,161,163,169] to enhance network reliability and resilience through optimal reconfiguration and topology enhancement in MDN.

7. Microgrids

7.1. From Planning Perspectives

Several concerns, such as economic, reliability, and environmental consequences, are always of great concern in the planning of MGs. On the one hand, optimal MG planning has a significant influence on and is directly linked with MG operation [170]. Reasonable planning ensures the optimal operation of MG. A substantial amount of research has been carried out in the literature. To investigate the economic sustainability of MG deployment, [171] developed a model for the MG planning issue under different uncertainties, such as unknown physical and financial information. In [172], a planning problem for provisional MG is developed and formulated. Different single-objective and multiobjective computational optimization techniques, such as MILP, MIP, SQP, DP, MINLP, IP, LP, MPDP, Lagrange multipliers KKT conditions, reduced gradient method, NSGA-II, simulation accounting, game theory, EA, GA, SA, PSO, AIS, Vaccine-AIS, MDPSO, MADS, MGSA, AMFA, GSA, SCSS, BFA, and CHASE, have been reported for optimal planning of MG from different aspects [173]. Other studies have been carried out on solving planning problems with the objective of minimizing different planning costs of MG [174,175,176]. Based on a semiempirical approach, an extensive roadmap is provided in the context of MG planning as a systemic solution that contributes to ensuring sustainability and resilience in terms of MG project M&O [177]. An extensive review of MG planning is presented in [170]. Microgrid design framework flow is given in Figure 19. The reviewed work for MGs planning perspective is summarized in Table 4.

Table 4. Microgrids: Reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Modes of Operation	Usage
[178]	Generation and load forecasting	1. (⇊) Scheduling cost 2. (⇈) System benefits	1. Load forecast 2. State estimation 3. UC 4. OPF	MG at Lab. for CER, School of EE Eng. in NTU	Short-term planning	ANN	--	--	Techno-economic	--	--
[179,180,181]	Generation and load forecasting	1. (⇊) Forecasting error 2. Load forecasting	1. Mean absolute percentage error 2. Load data 3. Internal and external inputs 4. Ramp-up and ramp-down limits	1. MG by Spanish com. Iberdrola 2. Real MG (Huatacondo, in northern Chile 3. Con. BCIT Vancouver, BC, Canada as MG	Short-term planning	1. ANN 2. SOM 3.SRWNN	-- -- --	-- -- --	Technical	-- -- --	-- -- --
[182]	Generation and load forecasting	1. (⇈) Power network control and management	1. Wind power forecasting 2. Resource scheduling 3. Wind speed 4. Wind power	Goldwing microgrid wind turbine system (Beijing, China)	Short-term planning	ANN	Wind Energy	--	Technical	--	--
[183,184]	Generation and load forecasting	1. (⇈) Power network control and management 2. (⇊) Forecasting error	1. Solar energy forecasting 2. Solar radiation forecasting 3. Weather data 4. Solar irradiance	Trained data from Euskalmet	1. Short-term planning 2. Long-term planning	1. ANN 2. DL	Solar PV	--	Technical	--	--
[185,186,187]	Generation and load forecasting	1. (⇈) Accuracy of electricity load forecasting	1. Load forecasting 2. Weather conditions 3. Electricity consumption	1. LV-MG of commercial bank 2. Rural MG in Africa 3. MG, rural Sub-Saharan Africa	Short-term planning	1. ANN 2. SVR-LSTM 3. DL and Bi-LSTM	1. PV Diesel gen. Storage -- --	1. Battery -- --	Technical	Grid-connected -- --	1. Commercial -- 2. Residential and commercial
[188]	Generation and load forecasting	1. (⇈) Economic benefits 2. (⇈) Environmental benefits	1. Load forecasting 2. Generation and load equality 3. Battery constr.	Exemplary MG	Multiobjective planning	MAS and PSO	Wind PV MT FC Storage	Battery Hydrogen	Econo-environment	Grid-connected	Residential
[189]	Generation and load forecasting	1. (⇊) Daily average forecast errors	1. Load forecast 2. Historical load demand dataset	MG system in Beijing, China	Short-term planning	Integration of GA, PSO, and ANFI- systems	WT PV MT Storage	Battery	Technical	Grid-connected	Industrial
[190]	Resource allocation (siting, sizing, control, and integration)	1. (⇊) Power exchange with the utility 2. (⇊) Duration of service interruption 3. (⇈) Reliability	1. T/F constraints 2. DER constraint 3. Storage constr. 4. Utility constr. 5. Operational constraints	SMG Testbed at BCIT, Burnaby, BC, Canada	Operational planning	FDISR algorithm	CHP-MT Storage	Battery	Techno-economic	Both grid-connected and islanding modes	Commercial (institutional)
[191,192]	Resource allocation (siting and sizing)	1. (⇊) Energy loss 2. (⇊) Power imbalance 3. (⇊) Cost	1. Penetration level 2. Power flow 3. Voltage limits 4. Power line 5. Current limits	69-bus distribution system	Short-term (1-year) planning	Graph portioning (GP) and tabu search (TS)	PV WT Biomass DRS	--	Techno-economic	Stand-alone	Residential
[193]	Resource allocation (siting and sizing)	1. (⇊) Voltage drops 2. (⇊) Line losses 3. (⇈) Energy savings	1. P and Q limits 2. Voltage limits 3. Current limits 4. Main grid constraints	Exemplary system (10 kV DG distribution circuit model)	Operational planning	Combined graph-based and GA	Different DGs Storage	Battery	Techno-economic	Grid-connected	Distribution (residential)
[194,195]	Resource allocation (siting and sizing)	1. (⇊) Operational and investment cost 2. (⇈) Profit	1. Power balance 2. Generation output limits 3. Grid constraint 4. Battery const.	1. Exemplary LV-MG 2. IEEE 37-bus standard network	1. Operation planning 2. Operation planning	1. MRCGA 2. HAS	PV FC MT Storage	Battery	Economic	1. Grid-connected	Distribution (residential)
[196]	Resource allocation (siting and sizing)	1. (⇊) Investment cost 2. (⇊) Expected energy not supplied 3. (⇊) Line losses	1. Power flow 2. Load capacity 3. DG and ES capacity limits 4. Thermal limits 5. DG gen. limits 6. Bus voltages 7. DG and ES gro.	IEEE 33-bus and real implementation of State Grid Corporation of China (SGCC)	Multiobjective planning	Improved NSGA-II algorithm	PV Wind Storage	Battery	Techno-economic	Both grid-connected and islanding modes	Commercial and distribution
[197]	Resource allocation (siting and sizing)	1. (⇊) Energy loss 2. (⇊) Energy cost	1. Power flow 2. Voltages limits 3. Demand covering 4. Battery constr.	31-bus MV-MG	Hourly planning	SQP	PV WT Storage	Battery	Techno-economic	Both grid-connected and stand-alone	Distribution (residential)
[198]	Resource allocation (economic dispatch and siting, sizing)	1. (⇊) Daily fixed cost of investment 2. (⇊) Load loss probability 3. (⇊) Excess energy rate 4. (⇊) O&M cost 5. (⇊) Pollutant disposal cost	1. Optimal allocation of DGs 2. Number of DGs 3. Storage system constraints 4. Weather conditions	Exemplary system	Multiobjective planning	Improved adaptive genetic algorithm (IAGA)	WT Diesel gen. PV Storage	Battery	Techno-economic and environment	Grid-connected	Distribution (residential)
[199]	Resource allocation (siting and sizing)	1. (⇊) Net present cost 2. (⇊) Purchased power from the grid	1. Voltage limits 2. Interrupting limits of critical and controllable loads 3. Exchange power limits	Exemplary MG system	Long-term (20 years) planning	PSO	PV WT MT Storage	Battery	Techno-economic	Both grid-connected and stand-alone	Distribution (residential)
[200]	Resource allocation (siting and sizing)	1. (⇈) Reliability 2. (⇈) PQ 3. (⇊) Power loss 4. (⇊) Cost	1. DR 2. Uncertainties 3. Load growth 4. Voltage limits 5. Frequency 6. Resources 7. ESSs	Ekbatan residential complex, Tehran, Iran	Multiobjective long-term (20 yrs.) planning	PSO	PV WT FC Storage	Battery Hydrogen	Techno-economic	Grid-connected	Residential
[201]	Resource allocation (type, siting, and sizing)	1. (⇊) Exchange power 2. (⇊) Cost 3. (⇊) Coalition payments	1. Voltage limits 2. Power cap. 3. Uncertainties 4. Prices 5. DG integration	33-MGs standard system	Multiobjective planning	CRPSO algorithm	PV WT	--	Techno-economic	Grid-connected	Industrial and commercial
[202,203,204,205,206]	Resource allocation (BESS siting, and sizing)	1. (⇊) Total cost 2. (⇊) Loss of load expectations 3. (⇊) Storage capacity	1. Nodal balance 2. N/W constr. 3. MT and REG constraints 4. BESS 5. Syst. reserve 6. FC constraint 7. Load curtail. and reliability 8. Interconnect. Constraints 9. DRP and TOU 10. Uncertainties	1. Modified IEEE 33-bus radial system 2. 33-bus MG 3. Test MG 4. MG in Northwest China 5. Modified IEEE 33-bus radial system	1. Long term 2. Short term (1 yr.) 3. Long term 4. Long term 5. Long-term planning	1. Backward scenario reduction method 2. ε-constraint method 3. MILP 4. PSO 5. HHO algorithm	1. PV, FC, MT, ESS 2. WT, diesel gen., ESS 3. Dispatch DGs, ESS 4. WT, PV, ESS 5. PV, WT, ESS	1. Battery, Hydrogen 2. Battery 3. Battery 4. Battery 5. Battery	1. Economic 2. Techno-economic 3. Economic 4. Technical 5. Economic	1. Both 2. Grid-connected 3. Both 4. Both 5. Both	1. Commercial and industrial 2. Residential 3. Residential and commercial 4. All 5. All
[207,208,209,210]	Resource allocation (D-STATCOM, switches, protection devices, and relay siting, and sizing)	1. (⇈) Voltage profile 2. (⇊) Unsupplied load 3. (⇊) No. of installed switches 4. (⇈) Reliability 5. (⇊) ENS	1. DSTATCOM 2. Feeder constraints 3. Voltage limits 4. Voltage angle 5. Switches 6. P&Q limits 7. Grid-forming DGs 8. Grid-following DGs	1. Exemplary system with 3 feeders 2. IEEE 13-bus standard 54-bus Italian DS 3. 25-bus syst. 4. Thailand MG-DS 5. 18-bus MG and IEEE 123-bus test system	1. Operation planning 2. Operation planning 3. Short-term planning 4. Short-term planning	1. New technique 2. Alliance algorithm 3. PSO and MCS 4. BMO-PSO 5. Exchange market algorithm	1. D-STATCOM 2. Switches 3. Protect. Devices 4. Relays	--	1. Technical 2. Technical 3. Technical 4. Relays	1. Both 2. Grid-connected 3. Both 4. Both	-- Residential Residential Residential
[211]	Topology selection	1. (⇈) Reliability 2. (⇈) Power supply 3. (⇈) Feasibility 4. (⇊) Power loss	1. Parallel DC/DC converter topology 2. Floating topology 3. 3-level neutral point clamped converter top.	Exemplary system	Operation planning	Comparative analysis	WT, PV, Diesel gen., ESS	Super capacitor, Battery	Technical	N/A	N/A
[212]	Topology selection	1. (⇈) Reliability 2. (⇊) Diff. b/w measured and calculated voltage angle and magnitude 3. (⇈) Accuracy	1. SCADA 2. Voltage magnitudes 3. Voltage angles 4. Phase angles 5. P and Q	Hypothetical MG test feeder	Operation planning	PMUs Voltage angle approach	--	--	Technical	N/A	N/A
[213,214]	Topology selection	1. (⇈) Reliability 2. (⇊) Inter-communications b/w adjacent parts 3. (⇊) Inter flows 4. (⇈) Gen/load balance	1. Loop based topology 2. Power balance 3. Voltage limits 4. Generation capacity	MG based on IEEE 37-bus feeder	Multiobjective operational planning	Graph partitioning and IP integrated method	DGs Energy storage facilities	Batteries	Technical	Both grid-connected and stand-alone	Industrial
[215,216]	Topology selection (Comm. Topology and controller)	1. (⇈) Converge performance 2. (⇊) Time delays 3. (⇊) Comm. cost	1. Controller gain 2. Volt. control 3. Freq. response 4. DG power rate 5. Droop coeff. 6. Bias factor 7. Line imped.	1. Test MG 2. 300 kW, 400 V DC-MG	Multiobjective planning	Graph theory and Consensus-based secondary voltage control	--	--	Techno-economic	Stand-alone	N/A
[217,218]	Topology selection (dispersed and centralized topologies)	1. (⇊) Power loss 2. (⇈) PV penetration	1. P&Q limits 2. Voltage limits 3. Voltage angles 4. Penetration level 5. Matching factor	MG of isolated community of Brazilian Amazon region	Short-term (1 year) planning	Graph theory	PV, Diesel, storage	Battery	Technical	Stand-alone	Residential and commercial

Table 4. Microgrids: Reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Modes of Operation	Usage
[178]	Generation and load forecasting	1. (⇊) Scheduling cost 2. (⇈) System benefits	1. Load forecast 2. State estimation 3. UC 4. OPF	MG at Lab. for CER, School of EE Eng. in NTU	Short-term planning	ANN	--	--	Techno-economic	--	--
[179,180,181]	Generation and load forecasting	1. (⇊) Forecasting error 2. Load forecasting	1. Mean absolute percentage error 2. Load data 3. Internal and external inputs 4. Ramp-up and ramp-down limits	1. MG by Spanish com. Iberdrola 2. Real MG (Huatacondo, in northern Chile 3. Con. BCIT Vancouver, BC, Canada as MG	Short-term planning	1. ANN 2. SOM 3.SRWNN	-- -- --	-- -- --	Technical	-- -- --	-- -- --
[182]	Generation and load forecasting	1. (⇈) Power network control and management	1. Wind power forecasting 2. Resource scheduling 3. Wind speed 4. Wind power	Goldwing microgrid wind turbine system (Beijing, China)	Short-term planning	ANN	Wind Energy	--	Technical	--	--
[183,184]	Generation and load forecasting	1. (⇈) Power network control and management 2. (⇊) Forecasting error	1. Solar energy forecasting 2. Solar radiation forecasting 3. Weather data 4. Solar irradiance	Trained data from Euskalmet	1. Short-term planning 2. Long-term planning	1. ANN 2. DL	Solar PV	--	Technical	--	--
[185,186,187]	Generation and load forecasting	1. (⇈) Accuracy of electricity load forecasting	1. Load forecasting 2. Weather conditions 3. Electricity consumption	1. LV-MG of commercial bank 2. Rural MG in Africa 3. MG, rural Sub-Saharan Africa	Short-term planning	1. ANN 2. SVR-LSTM 3. DL and Bi-LSTM	1. PV Diesel gen. Storage -- --	1. Battery -- --	Technical	Grid-connected -- --	1. Commercial -- 2. Residential and commercial
[188]	Generation and load forecasting	1. (⇈) Economic benefits 2. (⇈) Environmental benefits	1. Load forecasting 2. Generation and load equality 3. Battery constr.	Exemplary MG	Multiobjective planning	MAS and PSO	Wind PV MT FC Storage	Battery Hydrogen	Econo-environment	Grid-connected	Residential
[189]	Generation and load forecasting	1. (⇊) Daily average forecast errors	1. Load forecast 2. Historical load demand dataset	MG system in Beijing, China	Short-term planning	Integration of GA, PSO, and ANFI- systems	WT PV MT Storage	Battery	Technical	Grid-connected	Industrial
[190]	Resource allocation (siting, sizing, control, and integration)	1. (⇊) Power exchange with the utility 2. (⇊) Duration of service interruption 3. (⇈) Reliability	1. T/F constraints 2. DER constraint 3. Storage constr. 4. Utility constr. 5. Operational constraints	SMG Testbed at BCIT, Burnaby, BC, Canada	Operational planning	FDISR algorithm	CHP-MT Storage	Battery	Techno-economic	Both grid-connected and islanding modes	Commercial (institutional)
[191,192]	Resource allocation (siting and sizing)	1. (⇊) Energy loss 2. (⇊) Power imbalance 3. (⇊) Cost	1. Penetration level 2. Power flow 3. Voltage limits 4. Power line 5. Current limits	69-bus distribution system	Short-term (1-year) planning	Graph portioning (GP) and tabu search (TS)	PV WT Biomass DRS	--	Techno-economic	Stand-alone	Residential
[193]	Resource allocation (siting and sizing)	1. (⇊) Voltage drops 2. (⇊) Line losses 3. (⇈) Energy savings	1. P and Q limits 2. Voltage limits 3. Current limits 4. Main grid constraints	Exemplary system (10 kV DG distribution circuit model)	Operational planning	Combined graph-based and GA	Different DGs Storage	Battery	Techno-economic	Grid-connected	Distribution (residential)
[194,195]	Resource allocation (siting and sizing)	1. (⇊) Operational and investment cost 2. (⇈) Profit	1. Power balance 2. Generation output limits 3. Grid constraint 4. Battery const.	1. Exemplary LV-MG 2. IEEE 37-bus standard network	1. Operation planning 2. Operation planning	1. MRCGA 2. HAS	PV FC MT Storage	Battery	Economic	1. Grid-connected	Distribution (residential)
[196]	Resource allocation (siting and sizing)	1. (⇊) Investment cost 2. (⇊) Expected energy not supplied 3. (⇊) Line losses	1. Power flow 2. Load capacity 3. DG and ES capacity limits 4. Thermal limits 5. DG gen. limits 6. Bus voltages 7. DG and ES gro.	IEEE 33-bus and real implementation of State Grid Corporation of China (SGCC)	Multiobjective planning	Improved NSGA-II algorithm	PV Wind Storage	Battery	Techno-economic	Both grid-connected and islanding modes	Commercial and distribution
[197]	Resource allocation (siting and sizing)	1. (⇊) Energy loss 2. (⇊) Energy cost	1. Power flow 2. Voltages limits 3. Demand covering 4. Battery constr.	31-bus MV-MG	Hourly planning	SQP	PV WT Storage	Battery	Techno-economic	Both grid-connected and stand-alone	Distribution (residential)
[198]	Resource allocation (economic dispatch and siting, sizing)	1. (⇊) Daily fixed cost of investment 2. (⇊) Load loss probability 3. (⇊) Excess energy rate 4. (⇊) O&M cost 5. (⇊) Pollutant disposal cost	1. Optimal allocation of DGs 2. Number of DGs 3. Storage system constraints 4. Weather conditions	Exemplary system	Multiobjective planning	Improved adaptive genetic algorithm (IAGA)	WT Diesel gen. PV Storage	Battery	Techno-economic and environment	Grid-connected	Distribution (residential)
[199]	Resource allocation (siting and sizing)	1. (⇊) Net present cost 2. (⇊) Purchased power from the grid	1. Voltage limits 2. Interrupting limits of critical and controllable loads 3. Exchange power limits	Exemplary MG system	Long-term (20 years) planning	PSO	PV WT MT Storage	Battery	Techno-economic	Both grid-connected and stand-alone	Distribution (residential)
[200]	Resource allocation (siting and sizing)	1. (⇈) Reliability 2. (⇈) PQ 3. (⇊) Power loss 4. (⇊) Cost	1. DR 2. Uncertainties 3. Load growth 4. Voltage limits 5. Frequency 6. Resources 7. ESSs	Ekbatan residential complex, Tehran, Iran	Multiobjective long-term (20 yrs.) planning	PSO	PV WT FC Storage	Battery Hydrogen	Techno-economic	Grid-connected	Residential
[201]	Resource allocation (type, siting, and sizing)	1. (⇊) Exchange power 2. (⇊) Cost 3. (⇊) Coalition payments	1. Voltage limits 2. Power cap. 3. Uncertainties 4. Prices 5. DG integration	33-MGs standard system	Multiobjective planning	CRPSO algorithm	PV WT	--	Techno-economic	Grid-connected	Industrial and commercial
[202,203,204,205,206]	Resource allocation (BESS siting, and sizing)	1. (⇊) Total cost 2. (⇊) Loss of load expectations 3. (⇊) Storage capacity	1. Nodal balance 2. N/W constr. 3. MT and REG constraints 4. BESS 5. Syst. reserve 6. FC constraint 7. Load curtail. and reliability 8. Interconnect. Constraints 9. DRP and TOU 10. Uncertainties	1. Modified IEEE 33-bus radial system 2. 33-bus MG 3. Test MG 4. MG in Northwest China 5. Modified IEEE 33-bus radial system	1. Long term 2. Short term (1 yr.) 3. Long term 4. Long term 5. Long-term planning	1. Backward scenario reduction method 2. ε-constraint method 3. MILP 4. PSO 5. HHO algorithm	1. PV, FC, MT, ESS 2. WT, diesel gen., ESS 3. Dispatch DGs, ESS 4. WT, PV, ESS 5. PV, WT, ESS	1. Battery, Hydrogen 2. Battery 3. Battery 4. Battery 5. Battery	1. Economic 2. Techno-economic 3. Economic 4. Technical 5. Economic	1. Both 2. Grid-connected 3. Both 4. Both 5. Both	1. Commercial and industrial 2. Residential 3. Residential and commercial 4. All 5. All
[207,208,209,210]	Resource allocation (D-STATCOM, switches, protection devices, and relay siting, and sizing)	1. (⇈) Voltage profile 2. (⇊) Unsupplied load 3. (⇊) No. of installed switches 4. (⇈) Reliability 5. (⇊) ENS	1. DSTATCOM 2. Feeder constraints 3. Voltage limits 4. Voltage angle 5. Switches 6. P&Q limits 7. Grid-forming DGs 8. Grid-following DGs	1. Exemplary system with 3 feeders 2. IEEE 13-bus standard 54-bus Italian DS 3. 25-bus syst. 4. Thailand MG-DS 5. 18-bus MG and IEEE 123-bus test system	1. Operation planning 2. Operation planning 3. Short-term planning 4. Short-term planning	1. New technique 2. Alliance algorithm 3. PSO and MCS 4. BMO-PSO 5. Exchange market algorithm	1. D-STATCOM 2. Switches 3. Protect. Devices 4. Relays	--	1. Technical 2. Technical 3. Technical 4. Relays	1. Both 2. Grid-connected 3. Both 4. Both	-- Residential Residential Residential
[211]	Topology selection	1. (⇈) Reliability 2. (⇈) Power supply 3. (⇈) Feasibility 4. (⇊) Power loss	1. Parallel DC/DC converter topology 2. Floating topology 3. 3-level neutral point clamped converter top.	Exemplary system	Operation planning	Comparative analysis	WT, PV, Diesel gen., ESS	Super capacitor, Battery	Technical	N/A	N/A
[212]	Topology selection	1. (⇈) Reliability 2. (⇊) Diff. b/w measured and calculated voltage angle and magnitude 3. (⇈) Accuracy	1. SCADA 2. Voltage magnitudes 3. Voltage angles 4. Phase angles 5. P and Q	Hypothetical MG test feeder	Operation planning	PMUs Voltage angle approach	--	--	Technical	N/A	N/A
[213,214]	Topology selection	1. (⇈) Reliability 2. (⇊) Inter-communications b/w adjacent parts 3. (⇊) Inter flows 4. (⇈) Gen/load balance	1. Loop based topology 2. Power balance 3. Voltage limits 4. Generation capacity	MG based on IEEE 37-bus feeder	Multiobjective operational planning	Graph partitioning and IP integrated method	DGs Energy storage facilities	Batteries	Technical	Both grid-connected and stand-alone	Industrial
[215,216]	Topology selection (Comm. Topology and controller)	1. (⇈) Converge performance 2. (⇊) Time delays 3. (⇊) Comm. cost	1. Controller gain 2. Volt. control 3. Freq. response 4. DG power rate 5. Droop coeff. 6. Bias factor 7. Line imped.	1. Test MG 2. 300 kW, 400 V DC-MG	Multiobjective planning	Graph theory and Consensus-based secondary voltage control	--	--	Techno-economic	Stand-alone	N/A
[217,218]	Topology selection (dispersed and centralized topologies)	1. (⇊) Power loss 2. (⇈) PV penetration	1. P&Q limits 2. Voltage limits 3. Voltage angles 4. Penetration level 5. Matching factor	MG of isolated community of Brazilian Amazon region	Short-term (1 year) planning	Graph theory	PV, Diesel, storage	Battery	Technical	Stand-alone	Residential and commercial

Figure 19. A flowchart for microgrid design framework (adapted from [219]).

Figure 19. A flowchart for microgrid design framework (adapted from [219]).

7.1.1. Generation and Load Forecasting

Electricity demand forecasting is a critical and important part of MG operations and expansion plans. Because of the deregulation of energy markets, demand patterns are becoming increasingly complicated. As a result, artificial intelligence (AI)-based forecasting techniques are becoming increasingly popular for short- and long-term generation and load forecasting with low error margins. Researchers have investigated different AI-based techniques, such as ANN, deep neural networks (DNN), and ML approaches, for short-term generation (solar and wind) and load forecasting in the microgrid paradigm [178,179,180,181,182,183,185,186,187,220]. The authors in [188,189] proposed an MAS and binary genetic algorithm (BGA)-based short-term forecasting load and generation model for MG. From long-term forecasting perspectives, [184] proposed a deep learning model for solar radiation forecasting for installation of MGs.

7.1.2. Resource Allocation

Optimal allocation and construction of MGs includes optimal siting and sizing of DGs, ESSs and switches, breakers, and reactive power compensation devices within a network. Optimal allocation of MG resources enhances the resilience and reliability of MG and provides a cost-efficient solution for energy delivery to end users. Different approaches, such as chaotic multiobjective genetic algorithm (CMO-GA) [221], fault detection, isolation, and service restoration (FDISR) algorithm [190], graph partitioning and tabu search (GP&TS) [191,192], life cycle cost theory [222], combine graph-based and GA [193], matrix real-coded genetic algorithm (MRCGA) [194], harmony search algorithm (HSA) [195], NSGA-II [196], sequential quadratic programming (SQP) [197], hybrid weighted method (HWM) [198], PSO [199,200], coalition game theory (CGT) [201], and Exhaustive search algorithm (ESA) have been reported in the literature for optimal planning of DGs in MG framework. Similarly, Authors proposed backward scenario reduction (BSR), e-constraint method, MILP model, Improved PSO, Harris Hawks optimization (HHO), and structure preserving energy function algorithm based approaches for optimal siting and sizing of BESs for different applications in MG [202,203,204,205,206,223]. To tackle the losses at optimum, it is necessary to optimally locate the compensation device within MG framework. The authors in [224] proposed a distributed control strategy for optimal allocation of reactive power compensation to system. The authors in [207] proposed a planning and operational approach for optimal allocation of DSTATCOM for reactive power compensation in single phase operation of MGs. A state-of-the art review has been carried out to investigate the different reactive power compensation techniques in MGs [225]. While optimal placement of various switching and protection devices has been carried out in the literature through the alliance algorithm [208], PSO [209], multiobjective PSO [53], and exchange market algorithm (EMA) [210].

7.1.3. Topology Selection

The traditional distribution network is evolving into more interactive networks that can accommodate many DGs, ESSs, EVs, CPLs, and RESs. DC-MGs or hybrid DC/AC MGs interact inside distribution networks to alleviate several difficulties associated with traditional DN. However, the topological design of MGs is critical, posing a significant planning challenge for planning engineers [170]. Because most switches lack measuring equipment and communication links, the topology of DNs is frequently unclear. For safe and dependable grid operation, however, knowledge of the real topology is essential [212]. Topological design is a significant concern in MG planning for assuring specific features, such as high reliability in islanded operation. The authors in [226] developed a new DC-DC converter topology for microgrid applications. The authors in [211] compared three different topologies for hybrid ESS for MG applications. The authors in [212] proposed a novel voting-based approach for topology detection based on a micro-synchrophasor for the MG framework. By considering DERs, the authors in [213] proposed graph partitioning and integer programming integrated methodology for optimal planning of loop-based MGs. The authors in [214] suggested an iterative approach for optimum topology design for MGs in an ADN based on graph partitioning, integer programming, and performance index. The authors in [215,216] developed a communication topology for secondary distributed voltage control in islanded MGs and DC-MGs, respectively. From a protection perspective, the authors in [217] developed a model for automatic topology extraction of network based on graph portioning theory. Finally, [218] presented a comprehensive comparative analysis of distributed vs. centralized topologies in an autonomous MG for distant communities in the Brazilian Amazon region.

8. Multi-Microgrids

8.1. From Planning Perspectives

Planning engineers have put in a lot of effort to ensure MMG planning in an optimal way. Network reliability and efficacy are improved through optimal planning, increasing system reliability, deferring system upgrades, and lowering energy and power losses. Ref. [227] proposed a decentralized collaborative dispatch architecture of IMMG based on a multiagent consensus algorithm (MACA) for efficiently dispatching dispatchable active power resources and distributing total power command among all MGs in an MMG paradigm. By applying the imperialist competitive algorithm (ICA), [43] presented a systematic approach for optimal planning of MMGs by considering the significance of reliability with the aim of lowering the cost of electricity generation and voltage profile. Ref. [228] presented an optimal planning scheme to design CHP system with cooperative energy management in MMGs with the objectives of integrated hubs with lowest feasible cost, meeting demand, reducing GHG emissions, and creating a bidirectional interaction between electricity and gas networks. Optimal planning process for MG and MMG is shown in Figure 20. The reviewed work for MMG planning perspective is summarized in Table 5.

Table 5. Multi-microgrids: reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Modes of Operation	Usage
[43,227,228,229]	MMG planning	1. (⇊) Annual cost 2. (⇊) Total regulation cost 3. (⇈) Reliability 4. (⇈) Voltage profile 5. (⇊) Total ENS 6. (⇊) Emission	1. Power balance 2. Power out limit 3. Power exch. 4. Voltage limits 5. Installed cap. 6. Uncertainties 7. Line capacity 8. No. of MGs 9. On/off constr. 10. Technical con. 11. Penalty factor	1 and 2. Hybrid AC/DC-MG test system 3. Standard 69-bus DS 4. IEEE 6-bus test system	Operation planning	1. Bender’s decomposition method (BDM) 2. Multiagent consensus algorithm 3. ICA	1. PV, EV, WT, DG, Grid 2. MT, WT, PV, Grid, ES 3. WT, PV, Biomass 4. RG, DG, FC, ES, CHP, CCHP	-- 2. Battery -- 4. Battery and hydrogen	1. Economic 2. Economic 3. Techno-economic 4. Techno-econo-environment	1. Grid-connected 2. Grid-connected 3. Stand-alone 4. Stand-alone	Commercial and residential
[230,231,232]	Resource allocation (DGs and DPCC siting and sizing)	1. (⇈) Annual profit 2. (⇊) Total cost 3. (⇊) Power loss 4. (⇊) Emission	1. Power balance 2. Power out limit 3. Power exch. 4. V&I limits 5. Installed cap. 6. No. of MGs 7. Uncertainties	1. Mount Magnet, Western Australia 2. 85-bus distribution system 3. Modified IEEE 33-bus network	1. Short-term (1 year) planning 2. Multiobj planning	1.CGT 2. SMO-PSO 3. BOA and GOA	1. WT, PV, ES 2. WT, PV, CHP, DRS, ESS --	1. Battery 2. Battery --	1. Economic 2. Techno-econo-environment 3. Techno-economic	1. Grid-connected 2. Grid-connected 3. Grid-connected	Commercial and residential
[233]	Resource allocation	1. (⇊) Total cost 2. (⇊) EENS	1. No. of MGs 2. DR 3. N-1 contingency 4. Gen. constraint 5. Uncertainties	IEEE 123-bus test system	Short-term (daily) planning	ARO/MILP	DG ESS Capacitors	Battery	Economic	Both	Commercial, industrial, and residential
[234]	Resource allocation	1. (⇊) Total cost 2. (⇊) Prediction error	1. Uncertainties 2. RT-EM 3. DER capacity 4. Storage const. 5. Power exch.	MMG case of Ekbatan residential complex	Long-term planning	CM, CE, and game theory	WT, FC, EL, PV, EV, ES	Battery Hydrogen	Economic	Grid-connected	Residential
[230]	Resource allocation	1. (⇊) Total cost 2. (⇈) Profit	1. DERs output capacities 2. Converter topology 3. Battery constr.	MMG-town Mount Magnet in Western Australia	Short-term (1 year) planning	Cooperative game theory	WT PV ESS	Battery	Economic	Grid-connected	Residential
[235,236,237]	Resource allocation	1. (⇊) Loss 2. (⇊) Energy purchase from grid 3. (⇊) Emission 4. (⇈) Volt. profile 5. (⇈) Reliability 6. (⇈) Density 7. (⇊) Cost	1. Uncertainties 2. P and Q limit 3. Power factor 4. Load limits 5. Voltage limits 6. No. of DGs 7. Power balance	1. 69-bus RDS and modified 119-bus DS 3. Modified 33-bus DS	Multiobjective planning	1. Sensitivity analysis 2. Game theory 3. JAYA algorithm	1. PV WT Biomass 2. PV, WT, ESS 3. Diesel generator	-- Battery --	Techno-econo-environment	Grid-connected	Residential, industrial, and commercial
[238,239]	Topology selection	1. (⇈) Reliability 2. (⇈) Renewable penetration 3. (⇊) Number of SMGs on each cluster	1. No of MGs 2. Energy utilization 3. ILP formulation 4. DERs capacity 5. Storage const.	Exemplary system	Operation planning	Overlay topology design	MT PV CHP ESS	Battery	Technical	Both	Residential

Table 5. Multi-microgrids: reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Modes of Operation	Usage
[43,227,228,229]	MMG planning	1. (⇊) Annual cost 2. (⇊) Total regulation cost 3. (⇈) Reliability 4. (⇈) Voltage profile 5. (⇊) Total ENS 6. (⇊) Emission	1. Power balance 2. Power out limit 3. Power exch. 4. Voltage limits 5. Installed cap. 6. Uncertainties 7. Line capacity 8. No. of MGs 9. On/off constr. 10. Technical con. 11. Penalty factor	1 and 2. Hybrid AC/DC-MG test system 3. Standard 69-bus DS 4. IEEE 6-bus test system	Operation planning	1. Bender’s decomposition method (BDM) 2. Multiagent consensus algorithm 3. ICA	1. PV, EV, WT, DG, Grid 2. MT, WT, PV, Grid, ES 3. WT, PV, Biomass 4. RG, DG, FC, ES, CHP, CCHP	-- 2. Battery -- 4. Battery and hydrogen	1. Economic 2. Economic 3. Techno-economic 4. Techno-econo-environment	1. Grid-connected 2. Grid-connected 3. Stand-alone 4. Stand-alone	Commercial and residential
[230,231,232]	Resource allocation (DGs and DPCC siting and sizing)	1. (⇈) Annual profit 2. (⇊) Total cost 3. (⇊) Power loss 4. (⇊) Emission	1. Power balance 2. Power out limit 3. Power exch. 4. V&I limits 5. Installed cap. 6. No. of MGs 7. Uncertainties	1. Mount Magnet, Western Australia 2. 85-bus distribution system 3. Modified IEEE 33-bus network	1. Short-term (1 year) planning 2. Multiobj planning	1.CGT 2. SMO-PSO 3. BOA and GOA	1. WT, PV, ES 2. WT, PV, CHP, DRS, ESS --	1. Battery 2. Battery --	1. Economic 2. Techno-econo-environment 3. Techno-economic	1. Grid-connected 2. Grid-connected 3. Grid-connected	Commercial and residential
[233]	Resource allocation	1. (⇊) Total cost 2. (⇊) EENS	1. No. of MGs 2. DR 3. N-1 contingency 4. Gen. constraint 5. Uncertainties	IEEE 123-bus test system	Short-term (daily) planning	ARO/MILP	DG ESS Capacitors	Battery	Economic	Both	Commercial, industrial, and residential
[234]	Resource allocation	1. (⇊) Total cost 2. (⇊) Prediction error	1. Uncertainties 2. RT-EM 3. DER capacity 4. Storage const. 5. Power exch.	MMG case of Ekbatan residential complex	Long-term planning	CM, CE, and game theory	WT, FC, EL, PV, EV, ES	Battery Hydrogen	Economic	Grid-connected	Residential
[230]	Resource allocation	1. (⇊) Total cost 2. (⇈) Profit	1. DERs output capacities 2. Converter topology 3. Battery constr.	MMG-town Mount Magnet in Western Australia	Short-term (1 year) planning	Cooperative game theory	WT PV ESS	Battery	Economic	Grid-connected	Residential
[235,236,237]	Resource allocation	1. (⇊) Loss 2. (⇊) Energy purchase from grid 3. (⇊) Emission 4. (⇈) Volt. profile 5. (⇈) Reliability 6. (⇈) Density 7. (⇊) Cost	1. Uncertainties 2. P and Q limit 3. Power factor 4. Load limits 5. Voltage limits 6. No. of DGs 7. Power balance	1. 69-bus RDS and modified 119-bus DS 3. Modified 33-bus DS	Multiobjective planning	1. Sensitivity analysis 2. Game theory 3. JAYA algorithm	1. PV WT Biomass 2. PV, WT, ESS 3. Diesel generator	-- Battery --	Techno-econo-environment	Grid-connected	Residential, industrial, and commercial
[238,239]	Topology selection	1. (⇈) Reliability 2. (⇈) Renewable penetration 3. (⇊) Number of SMGs on each cluster	1. No of MGs 2. Energy utilization 3. ILP formulation 4. DERs capacity 5. Storage const.	Exemplary system	Operation planning	Overlay topology design	MT PV CHP ESS	Battery	Technical	Both	Residential

Figure 20. Optimal planning process of microgrid and multi-microgrid structures (adapted from [240]).

Figure 20. Optimal planning process of microgrid and multi-microgrid structures (adapted from [240]).

8.1.1. Generation and Load Forecasting

MMG performance is influenced by the kind of sources and ESS, capacity allocation, and other aspects of optimum design. As a result, the single MG concept does not completely apply to MMG planning and design issues. For planning, these to be thoroughly assessed, must consider demand and feasibility analysis, configuration optimization, and electrical design, benefit evaluation, and other important variables [44]. MMG planning is a complicated problem that considers various energy generating and storage technologies, as well as numerous energy vectors, in order to provide energy demands while minimizing I&O costs [229]. Therefore, a lot of research has been carried out to solve these types of planning issues in MMG structure in last decade. Various tools, such as REopt, RETScreen, SAM, HOMER, and DER-CAM, are reported in the literature to address MMG planning issues in terms of I&O cost minimization [229]. There is currently a critical need to solve optimal generation sources, storage, and resource allocation planning to complete various tasks and maintain the security, reliability, efficiency, and cost-effective operation of MMGs. Different approaches, such as the two-stage imperialist competitive algorithm (2S-ICA) [43], cooperative game theory (CGT) [230], and stochastic multiobjective PSO framework [241], have been reported in the literature for optimal resource allocation planning of MMGs based on network reliability. With the objectives of improving the voltage profile and reducing losses, [231] proposed a combined approach based on the butterfly and grasshopper optimization algorithm for optimal sizing and siting of distributed power controller. Optimization approaches were evaluated for coordinated operation and expansion planning problem of ADN with MMGs, DERs, DR, and N-1 generation contingency under different RES uncertainties [233,234]. A cooperative game theory approach was evaluated in [230] for optimal planning of clustered MGs in the presence of RESs (wind and solar) and storage systems. For optimal allocation of power command of island MMG across all MGs, a bilevel multiagent consensus algorithm (solved by GA)-based decentralized collaborative dispatch (DCD) approach is presented in [242]. Based on recent advances in artificial intelligence (AI) techniques, deep learning has opened new avenues for solving complex power system problems [243]. In [244], a novel data-driven method for DSM based on DNN and model-free reinforcement leaning (MFRL) is proposed with the aim of optimizing the power sale benefit with the upper network while minimizing the system’s peak to average rati

8.1.2. Resource Allocation

Optimal sizing and siting of different MMG components including DGs, storage systems and other reactive power compensation devices play a vital role for technical, economic, and environmentally friendly delivery of energy to end users. The authors in [235,236,237,245] proposed multiobjective approaches for optimal sizing and siting of DGs for construction of MMG based on sensitivity analysis, Nash equilibrium game theory, and the Jaya algorithm. Similarly, the authors in [228,246,247] performed an optimal sizing of ESSs with objectives of cost minimization and maximization of efficiency through NSGA-II, non-cooperative and cooperative optimization approaches. Effective energy transfer, as a result of active management and coordination in the MMG paradigm, obviously assists in enhancing the overall network’s stability and reliability [248]. Proper energy management (EM) greatly facilitates the system’s economics, flexibility and other performance measures, such as environmental concerns and system stability. MAS is a popular EM technique for large systems including several types of interacting players [248]. Various other EM strategies reported in the literature include dual decomposition, contingency-based EM, cooperative model predictive control (MPC), hierarchical power scheduling, outage management scheme and MPC, multiagent-based hierarchal EM strategy, alternating direction method of multipliers, deep reinforcement learning, and the tabu search algorithm [249,250,251,252,253,254,255,256]. The authors of [232] developed a stochastic multiobjective framework for optimum EM in the presence of DRP, which reduced carbon emissions by 13.4% and 9.2%, respectively, and increased MMG independence by 17% and 11.7% in case studies.

8.1.3. Topology Selection

Since an MG is a subset of a smart grid, a large cluster of different MGs leads to optimal smart grid design. Integration of ICT developments with power grid technologies is transforming the traditional grid’s architecture and operation, resulting in the grid’s progressive evolution into a smart grid. The smart grid’s generation will be different from today’s system, owing to the significant penetration of DEGs. To accommodate the renewable technologies in a better way and coupling of these MMGs with main power grids greatly depends upon the topology and structure of interconnected systems. Various research groups looked at different aspects of optimal topology design of interconnected structures of MGs. The authors in [238] designed an overlay reliable topology to enhance the usage of renewable energy in smart MG. To enhance the reliability and economic operation of MMGs, based on minimum cut-set methodology, the interconnection between network of MMGs was investigated and discussed in [239].

9. Virtual Power Plants

9.1. From Planning Perspectives

The authors of [257] utilized DSM to develop an optimum VPP planning model with the goal of reducing investment and operating expenses while taking into account the installation of wind power units and feeders. In [258], the authors proposed a composite system cost–worth analysis-based planning methodology for VPP aggregators to determine an appropriate DG portfolio with the goal of maximizing cost reductions in the day-ahead energy market. The authors in [259] proposed a scenario-based two-stage stochastic programming model to explore the expansion planning problem of a VPP trading energy in electricity markets. Ref. [260] suggested a probabilistic planning of VPP for participating in the energy and reserve markets using a MIP-based piece-wise model, with the goal of maximizing VPP profitability in DN. By considering the uncertainties of DERs and EVs, [261] proposed a MILP-based model for optimal planning of VPP. The VPP setup for district level from the concept of MMG is shown in Figure 21. The schematic of VPP is shown in Figure 22. The reviewed literature of VPP from planning perspective is shown in Table 6.

Table 6. Virtual power plants: reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Modes of Operation	Usage
[257]	VPP optimal planning	1. (⇊) Investment and operation cost 2. (⇊) Charging cost	1. DRP 2. EVs 3. Feeders constr. 4. WPP constr. 5. Gas turbines 6. Power balance 7. DN constraint	Modified 9-node DN	Long-term (3 years) planning	MINLP	WT EV Grid ESS	Battery	Economic	Grid-connected	Residential
[258]	VPP optimal planning	1. (⇈) VPP energy export 2. (⇊) Total energy curtailment 3. (⇊) Load curtailment and system generation cost	1. Network flow 2. DG curtailment limits 3. Line power flows limits 4. Generation limits 5. Thermal limits	IEEE Reliability Test System in conjunction with IEEE 69-node DN	Long-term planning	Composite System Cost/Worth Analysis	WT PV	--	Techno-economic	Grid-connected	Residential, commercial, and industrial
[260]	VPP optimal planning	1. (⇈) VPP profitability 2. (⇊) Demand cost	1. Thermal power plants 2. DG capacities 3. ESS constraint 4. Energy market prices	15-bus test system	Long-term planning	MIP and benders decomposition	TPP PV WT ESS	Electro-chemical	Economic	Stand-alone	Residential
[262]	Generation and load forecasting	1. (⇈) VPP profit in the market	1. Energy exch. 2. Capacity of interconnection 3. DA market 4. Flexible load 5. Production 6. Start-up and shut-down constr. 7. Min up/down constraints	18-bus distribution system	Short term and medium-term planning	MILP-ROM	WT PV	--	Economic	Grid-connected	Commercial VPP
[263]	Generation and load forecasting	1. (⇈) VPP profit in the market 2. (⇈) Conditional value at risk (CVaR)	1. Investment constraints 2. Power balance 3. CG operation 4. Flexible demand operation 5. REG constr. 6. ESS constraint 7. Risk metric constraints	Exemplary system	Long-term planning	Risk-constrained stochastic approach	PV ESS Convention generation	Battery	Techno-economic	Grid-connected	Residential, commercial, and industrial
[264]	Generation and load forecasting	1. (⇈) Expected day-ahead profit 2. (⇊) Expected day-ahead emissions	1. Uncertainties 2. Market price 3. Technical constraints 4. ESS constraint 5. Thermal limit 6. Power balance	Local network test system	Multiobjective planning	Two stage stochastic programming PSO/MOPSO	PV WT ESS CHP	Electro-chemical Battery	Econo-environment	Grid-connected	Commercial VPP
[265]	Generation and load forecasting	1. (⇊) Purchasing primary energy cost 2. (⇊) Fixed and variable generation cost 3. (⇊) Generation and operation cost 4. (⇊) Pollutant emission cost	1. Elect. demand 2. Cap. limit 3. Emission constraints 4. Peak demand 5. RE available 6. Dispatchable loads regulation 7. Elect. defici. 8. Resource availability 9. Nonnegativity constraints	Heterogenous test VPP system	Medium-term planning	Inexact two-stage stochastic linear programming (ITSLP) method	HP WT PV CHP ESS EVs	Battery	Econo-environment	Grid-connected	Commercial VPP
[266]	Generation and load forecasting	1. (⇈) Expected economic profit 2. (⇊) Deviation between actual energy output and contracted energy	1. Load balance 2. WT operation 3. DG constraint 4. IL constraint 5. ESS constraint 6. N/w constraint 7. Bid capacity 8. VPP reserve	10-bus test VPP structure	Short term operational planning	MPC-based dispatch model	WT DG ESS	Battery	Techno-economic	Grid-connected	Technical
[267,268,269]	Resource allocation (ESS sizing and siting)	1. (⇊) ESS cost 2. (⇊) Overall cost of VPP 3. (⇊) Power fluctuations 4. (⇊) Variance of output power 5. (⇊) Investment and maintenance cost	1. Uncertainties 2. C/D power constraints 3. Cap. Constr. 4. P and residual energy coupling 5. Power balance 6. Volt. security 7. Feeder security 8. P fluctuation 9. Capacity and rating 10. Curtailable load 11. Up/down limits	1. Modified IEEE-33 test feeder 2. 18-bus test system 3. Exemplary system	1. Long-term planning 2. Long-term (10 years) planning 3. Long-term (3 year) planning	1. Bilevel fuzzy stochastic expectation model (FSEM) and MCS, GA, QP 2. MINLP 3. Stochastic programming	1. PV, WT, ESS 2. WT, PV, ESS, Diesel generator 3. PV, WT, ESS	1. Battery 2. Battery 3. Battery	1. Techno-economic 2. Economic 3. Economic	1. Stand-alone 2. Grid-connected 3. Grid- connected	1. Commercial VPP 2. Commercial VPP 3. Commercial VPP
[270]	Resource allocation (Cascaded hydropower and PV’s sizing and siting)	1. (⇊) Overall cost 2. (⇊) Rate of change of total power output	1. Water quantity 2. Output power limits 3. RU/RD limits 4. No. of units 5. Upper/lower limits	Exemplary system	Medium-term and long-term planning	Auto regressive moving average model (ARIMA) model and Sensitivity analysis	HP TP PV	--	Techno-economic	Grid-connected	Commercial VPP
[271]	Resource allocation (DGs and EVs sizing and siting)	1. (⇈) Net income 2. (⇈) Clean energy consumption 3. (⇊) Emissions	1. Unit output and climbing 2. EV constraint 3. Controllable load constraints 4. ESS constraint 5. Grid interact 6. Scene probability cons. 7. Power balance	VPP in certain area of Shanxi Province	Multiobjective planning	CVaR	PV WT EVs	Battery	Techno-econo-environment	Grid-connected	Commercial VPP
[272]	Resource allocation (DGs and EVs sizing and siting)	1. (⇈) Incentive revenue of VPP operators 2. (⇈) System marginal price 3. (⇈) Renewable energy certificate market	1. Uncertainties 2. Number of DERs 3. Capacity of VPP 4. Charging and discharging power	Exemplary system	Multiobjective planning	Stochastic mixed-integer programming (SMIP)	WT PV ESS	Battery	Economic	Grid-connected	Commercial VPP
[273,274,275,276]	Energy management and Demand response	1. (⇈) Load restoration 2. (⇊) Load supply cost 3. (⇈) VPP profit 4. (⇊) VPP operating cost 5. (⇊) Losses 6. (⇈) Voltage profile	1. Demand resp. 3. Conn. Constr. 4. Slack, parent buses condition 5. Radial condit. 6. ESS constraint 7. Contr. load 8. Power balance 9. Energy exch. 10. Load control 11. N/w security 12. Power flow 13. EVs constr.	1. 118-bus test system 2. VPP in Siwa Oasis, Egypt. Siwa Oasis 3. Modified IEEE 13-bus test feeder 4. Exemplary VPP setup	Short term and medium-term planning	1. Stochastic MILP model 2. Electro-economical model 3. Risk-constrained stochastic programming 4. HOMER	1. WT ESS, EVs 2. FES, DEG, ESS, PV 3. CCHP, RES, HB, TS, ESS 4. PV, WT, EL, FC, HTU, ESS	1. Battery 2. Battery 3. Thermal and battery 4. Battery and hydrogen	1. Techno-economic 2. Techno-economic 3. Economic 4. Techno-economic	1. Grid-connected 2. Grid-connected 3. Grid-connected 4. Grid-connected	1. Commercial VPP 2. Commercial VPP 3. Commercial VPP 4. Commercial VPP
[277]	Topology selection	1. (⇈) Efficiency and reliability 5. (⇊) Losses 1. (⇈) Economic feasibility	1. VPP topology 2. Power limits 3. Lines capacity 4. Structural compactness 5. Power angle	MV-VPP test system of remote settlement power supply system	Operational planning	Graph theory and structural-topological analysis	WPP Gas piston unit PV	Battery	Techno-economic	Grid-connected	Commercial
[278]	Generation and load forecasting	1. (⇈) Forecasting accuracy 2. (⇊) Operation cost	1. Metrological conditions 2. Power balance 3. Power output limits 4. Ramp-up, ramp-down limit 5. Up and down time limits	Exemplary system	Short-term planning	Combination of EMD, CFNN, and LM	PV WT FC TPP	Hydrogen	Techno-economic	Stand-alone	Commercial
[279]	Generation and load forecasting	1. (⇈) PV integration 2. (⇊) Mean error	1. Solar irradiance forecasting 2. Irradiance 3. temperature 4. Weather data	Real PV installation test case	Short-term planning	ANN and novel similar hour-based selection (NSHS) algorithm	PV	--	Technical	--	Commercial
[280]	Generation and load forecasting	1. (⇈) Forecasting accuracy	1. Energy forecasting 2. Weather data 3. Consumption 4. Holidays 5. Random event 6. Flexible load 7. Historical data 8. Comfort index	Sydney/New South Wales (NSW) electricity grid	Short-term planning	ANN	--	--	Technical	--	Technical
[281]	Generation and load forecasting	1. (⇊) Forecasting error 2. (⇈) Revenue 3. (⇊) Operational cost of DERs and ESS	1. Forecast error 2. Units states 3. Non-anticipatory constraints	Exemplary system	Short-term planning	Two-stage stochastic optimization model	PV WT ESS	Battery	Techno-economic	Grid-connected	Commercial

Table 6. Virtual power plants: reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Modes of Operation	Usage
[257]	VPP optimal planning	1. (⇊) Investment and operation cost 2. (⇊) Charging cost	1. DRP 2. EVs 3. Feeders constr. 4. WPP constr. 5. Gas turbines 6. Power balance 7. DN constraint	Modified 9-node DN	Long-term (3 years) planning	MINLP	WT EV Grid ESS	Battery	Economic	Grid-connected	Residential
[258]	VPP optimal planning	1. (⇈) VPP energy export 2. (⇊) Total energy curtailment 3. (⇊) Load curtailment and system generation cost	1. Network flow 2. DG curtailment limits 3. Line power flows limits 4. Generation limits 5. Thermal limits	IEEE Reliability Test System in conjunction with IEEE 69-node DN	Long-term planning	Composite System Cost/Worth Analysis	WT PV	--	Techno-economic	Grid-connected	Residential, commercial, and industrial
[260]	VPP optimal planning	1. (⇈) VPP profitability 2. (⇊) Demand cost	1. Thermal power plants 2. DG capacities 3. ESS constraint 4. Energy market prices	15-bus test system	Long-term planning	MIP and benders decomposition	TPP PV WT ESS	Electro-chemical	Economic	Stand-alone	Residential
[262]	Generation and load forecasting	1. (⇈) VPP profit in the market	1. Energy exch. 2. Capacity of interconnection 3. DA market 4. Flexible load 5. Production 6. Start-up and shut-down constr. 7. Min up/down constraints	18-bus distribution system	Short term and medium-term planning	MILP-ROM	WT PV	--	Economic	Grid-connected	Commercial VPP
[263]	Generation and load forecasting	1. (⇈) VPP profit in the market 2. (⇈) Conditional value at risk (CVaR)	1. Investment constraints 2. Power balance 3. CG operation 4. Flexible demand operation 5. REG constr. 6. ESS constraint 7. Risk metric constraints	Exemplary system	Long-term planning	Risk-constrained stochastic approach	PV ESS Convention generation	Battery	Techno-economic	Grid-connected	Residential, commercial, and industrial
[264]	Generation and load forecasting	1. (⇈) Expected day-ahead profit 2. (⇊) Expected day-ahead emissions	1. Uncertainties 2. Market price 3. Technical constraints 4. ESS constraint 5. Thermal limit 6. Power balance	Local network test system	Multiobjective planning	Two stage stochastic programming PSO/MOPSO	PV WT ESS CHP	Electro-chemical Battery	Econo-environment	Grid-connected	Commercial VPP
[265]	Generation and load forecasting	1. (⇊) Purchasing primary energy cost 2. (⇊) Fixed and variable generation cost 3. (⇊) Generation and operation cost 4. (⇊) Pollutant emission cost	1. Elect. demand 2. Cap. limit 3. Emission constraints 4. Peak demand 5. RE available 6. Dispatchable loads regulation 7. Elect. defici. 8. Resource availability 9. Nonnegativity constraints	Heterogenous test VPP system	Medium-term planning	Inexact two-stage stochastic linear programming (ITSLP) method	HP WT PV CHP ESS EVs	Battery	Econo-environment	Grid-connected	Commercial VPP
[266]	Generation and load forecasting	1. (⇈) Expected economic profit 2. (⇊) Deviation between actual energy output and contracted energy	1. Load balance 2. WT operation 3. DG constraint 4. IL constraint 5. ESS constraint 6. N/w constraint 7. Bid capacity 8. VPP reserve	10-bus test VPP structure	Short term operational planning	MPC-based dispatch model	WT DG ESS	Battery	Techno-economic	Grid-connected	Technical
[267,268,269]	Resource allocation (ESS sizing and siting)	1. (⇊) ESS cost 2. (⇊) Overall cost of VPP 3. (⇊) Power fluctuations 4. (⇊) Variance of output power 5. (⇊) Investment and maintenance cost	1. Uncertainties 2. C/D power constraints 3. Cap. Constr. 4. P and residual energy coupling 5. Power balance 6. Volt. security 7. Feeder security 8. P fluctuation 9. Capacity and rating 10. Curtailable load 11. Up/down limits	1. Modified IEEE-33 test feeder 2. 18-bus test system 3. Exemplary system	1. Long-term planning 2. Long-term (10 years) planning 3. Long-term (3 year) planning	1. Bilevel fuzzy stochastic expectation model (FSEM) and MCS, GA, QP 2. MINLP 3. Stochastic programming	1. PV, WT, ESS 2. WT, PV, ESS, Diesel generator 3. PV, WT, ESS	1. Battery 2. Battery 3. Battery	1. Techno-economic 2. Economic 3. Economic	1. Stand-alone 2. Grid-connected 3. Grid- connected	1. Commercial VPP 2. Commercial VPP 3. Commercial VPP
[270]	Resource allocation (Cascaded hydropower and PV’s sizing and siting)	1. (⇊) Overall cost 2. (⇊) Rate of change of total power output	1. Water quantity 2. Output power limits 3. RU/RD limits 4. No. of units 5. Upper/lower limits	Exemplary system	Medium-term and long-term planning	Auto regressive moving average model (ARIMA) model and Sensitivity analysis	HP TP PV	--	Techno-economic	Grid-connected	Commercial VPP
[271]	Resource allocation (DGs and EVs sizing and siting)	1. (⇈) Net income 2. (⇈) Clean energy consumption 3. (⇊) Emissions	1. Unit output and climbing 2. EV constraint 3. Controllable load constraints 4. ESS constraint 5. Grid interact 6. Scene probability cons. 7. Power balance	VPP in certain area of Shanxi Province	Multiobjective planning	CVaR	PV WT EVs	Battery	Techno-econo-environment	Grid-connected	Commercial VPP
[272]	Resource allocation (DGs and EVs sizing and siting)	1. (⇈) Incentive revenue of VPP operators 2. (⇈) System marginal price 3. (⇈) Renewable energy certificate market	1. Uncertainties 2. Number of DERs 3. Capacity of VPP 4. Charging and discharging power	Exemplary system	Multiobjective planning	Stochastic mixed-integer programming (SMIP)	WT PV ESS	Battery	Economic	Grid-connected	Commercial VPP
[273,274,275,276]	Energy management and Demand response	1. (⇈) Load restoration 2. (⇊) Load supply cost 3. (⇈) VPP profit 4. (⇊) VPP operating cost 5. (⇊) Losses 6. (⇈) Voltage profile	1. Demand resp. 3. Conn. Constr. 4. Slack, parent buses condition 5. Radial condit. 6. ESS constraint 7. Contr. load 8. Power balance 9. Energy exch. 10. Load control 11. N/w security 12. Power flow 13. EVs constr.	1. 118-bus test system 2. VPP in Siwa Oasis, Egypt. Siwa Oasis 3. Modified IEEE 13-bus test feeder 4. Exemplary VPP setup	Short term and medium-term planning	1. Stochastic MILP model 2. Electro-economical model 3. Risk-constrained stochastic programming 4. HOMER	1. WT ESS, EVs 2. FES, DEG, ESS, PV 3. CCHP, RES, HB, TS, ESS 4. PV, WT, EL, FC, HTU, ESS	1. Battery 2. Battery 3. Thermal and battery 4. Battery and hydrogen	1. Techno-economic 2. Techno-economic 3. Economic 4. Techno-economic	1. Grid-connected 2. Grid-connected 3. Grid-connected 4. Grid-connected	1. Commercial VPP 2. Commercial VPP 3. Commercial VPP 4. Commercial VPP
[277]	Topology selection	1. (⇈) Efficiency and reliability 5. (⇊) Losses 1. (⇈) Economic feasibility	1. VPP topology 2. Power limits 3. Lines capacity 4. Structural compactness 5. Power angle	MV-VPP test system of remote settlement power supply system	Operational planning	Graph theory and structural-topological analysis	WPP Gas piston unit PV	Battery	Techno-economic	Grid-connected	Commercial
[278]	Generation and load forecasting	1. (⇈) Forecasting accuracy 2. (⇊) Operation cost	1. Metrological conditions 2. Power balance 3. Power output limits 4. Ramp-up, ramp-down limit 5. Up and down time limits	Exemplary system	Short-term planning	Combination of EMD, CFNN, and LM	PV WT FC TPP	Hydrogen	Techno-economic	Stand-alone	Commercial
[279]	Generation and load forecasting	1. (⇈) PV integration 2. (⇊) Mean error	1. Solar irradiance forecasting 2. Irradiance 3. temperature 4. Weather data	Real PV installation test case	Short-term planning	ANN and novel similar hour-based selection (NSHS) algorithm	PV	--	Technical	--	Commercial
[280]	Generation and load forecasting	1. (⇈) Forecasting accuracy	1. Energy forecasting 2. Weather data 3. Consumption 4. Holidays 5. Random event 6. Flexible load 7. Historical data 8. Comfort index	Sydney/New South Wales (NSW) electricity grid	Short-term planning	ANN	--	--	Technical	--	Technical
[281]	Generation and load forecasting	1. (⇊) Forecasting error 2. (⇈) Revenue 3. (⇊) Operational cost of DERs and ESS	1. Forecast error 2. Units states 3. Non-anticipatory constraints	Exemplary system	Short-term planning	Two-stage stochastic optimization model	PV WT ESS	Battery	Techno-economic	Grid-connected	Commercial

Figure 21. VPP setup at district level with respect to MMG concept.

Figure 21. VPP setup at district level with respect to MMG concept.

Figure 22. Schematic of virtual power plants (adapted from [282]).

Figure 22. Schematic of virtual power plants (adapted from [282]).

9.1.1. Generation and Load Forecasting

The process of defining goals, standards, and strategies for serving and delivering cost-effective power to active users is known as optimal planning and design of the VPP. The goal of VPP planning is to develop and analyze the technological and financial viability to find the least expensive approach with the greatest possible advantages. Planning objectives of VPP can be classified on the basis of environmental, economic and technical concerns [283]. Environmental objectives may include reduction in GHG emissions and maximization of RES penetration. Economic objectives cover maximization of overall profit and minimization of operational cost and network losses, while maximization of PQ and reliability and providing maintenance support service comes under the umbrella of technical objectives.

VPPs play an important role by participating in energy trading in the electricity market. Information technologies and advance metering infrastructure (AMI) are the backbone of VPP, which enable energy trading in both retail and wholesale markets. VPPs in an electricity market must first sign certain bilateral contracts before engaging in the day-ahead market, but the risk involved due to the intermittent nature of DGs can also affect these bilateral contracts. Therefore, efficient planning in VPPs is a crucial enabling tool for perfect operation from the economic and environmental points of view. With the aim of VPP profit maximization, a risk-bounded tool was developed in [262] to reduce the risk of not meeting bilaterally negotiated energies in an uncertain energy market environment due to a combination of heterogeneous DGs. The authors in [263] provided a unique technique for the investment planning of a VPP trading energy in an electricity market to attain an optimal anticipated return based on a nonconvex operational model (NOM). Evaluation of cost-effective and environmentally friendly planning solution is in high demand in power system research. The authors in [258,260,264,265] proposed different planning strategies, such as multiobjective day-ahead stochastic planning, MILP, inexact two-stage stochastic linear programming (ITSLP), and linear programming-based composite system cost–worth analysis for optimal planning of VPP from environmental end economic perspectives in the energy market. In [266], a two-stage MPC strategy-based operational planning framework was developed to optimize bids in the energy market with the goal of maximizing projected economic return. In [284], the authors proposed a power schedule planning and operation algorithm for efficiently distributing a defined power schedule to a large number of CHP plants and balancing variance in real-time operation in the VPP.

The production of RESs (wind and solar) and load demand at each time step during the day is very intermittent. Variation in RES production power and load demand are significant factors in the VPP’s economic functioning. When high levels of RES penetration and increased load fluctuation are predicted, effective forecasting of RE output and demand for VPP is required. In [278], a fusion method is used to combine empirical mode decomposition (EMD), cascade-forward neural network (CFNN), and linear model (LM) for adaptive load dispatching and forecasting in the presence of renewable hybrid systems. The authors in [285,286] and [287] proposed an MAS model, ML techniques, and long short-term memory network (LSTM) for short-term energy demand forecasting in VPP by considering DR. To enhance the penetration of RESs, [288] proposed load and generation forecasting-based decentralized control for dispatchable VPP. The intermittent nature of PV irradiance greatly effects the power production of VPP. The authors in [279] and [280] proposed ANN-based techniques for DA irradiance of PV and energy forecasting in VPP. An operational planning approach was developed by considering the forecasting error of DERs in VPP [281].

9.1.2. Resource Allocation

Optimal allocation of resources in VPP includes optimal sizing and siting of system assets to achieve higher reliability and cost-effective solutions. Investment in ESSs has a major impact on the economy of VPP. As a result, ESS is the most widely utilized tool in a VPP for adjusting power demand fluctuations to a given level of power output, so the role of ESS in VPP was investigated by [267,268], a bilevel fuzzy stochastic expectation model (FSEM) and risk-constrained stochastic model was evaluated for optimal allocation and to reduce the cost of storage in VPPs. VPP’s operational flexibility is enhanced through resource and asset planning, which reduces the economic risks it confronts in the short-term power market. By incorporating both DR and BESS, [269] proposed a data-driven approach for BESS sizing and DR load selection. Similarly, an autoregressive moving average model (ARIMA) model and complementary index [270], dynamic programming-based model [289], and CVaR method [271] were reported in the literature for optimal allocation of resources in a VPP framework. The authors in [272] proposed a stochastic MIP-based DER allocation method to maximize the VPP operator revenue by taking into account various market incentives.

VPP has a characteristic of local EMS and dynamic load, which might be a good option for DGs and network issues. VPP may communicate with DSO and each local EMS of each DG via main EMS to achieve optimal choice about when, which, and how much each DG should supply. Researchers have proposed a stochastic MILP model [273], electro-economic model [273], decentralized EM model [290], risk-constrained stochastic programming-based CVaR approach [291], and DR-based optimal planning [274] for energy management through DSM in the VPP paradigm under ESSs and a DR program. To address the security and reliability issues, [275] considered DG sizing and pricing and evaluated a point prediction method and PSO for optimal energy management of VPP with the aim of minimizing total operation cost and losses and maximizing the overall voltage profile.

9.1.3. Topology Selection

VPP’s optimal topology features assist in evaluating its reliability and efficiency. As a network with an optimum architecture, VPP offers users with high reliability and quality power supply at low prices and with minimum electrical losses. In [277], a graph theory (GT) and structural topological-based approach was evaluated for determining an optimal topology for VPP. The authors in [292] carried out an extensive review on structure and operation of VPP with respect to DER integration and operation, as well as VPP involvement in the energy market.

10. Smart Homes

10.1. From Planning Perspectives

The term “optimal SH planning” refers to the process of designing an SH that meets the requirements of inhabitants while being cost-effective [293]. The authors of [294] proposed an optimal multipath planning approach for wireless sensor and actor networks (WSANs) based on the Edmonds–Karp maximum flow algorithm (EKMFA) and Goldberg and Tarjan minimum-cost flow algorithm (GTMCFA) to set up paths to deliver intrusion event notification from event sources to actors in SHs. The authors in [295] developed an eco-trade bilinear algorithm to address the planning and scheduling problem by scheduling households’ loads, storage and energy sources among SHs in a community MG. The authors of [296] suggested a technique for evaluating and improving planning SHs based on knowledge-based rules using a hybrid multiple-attribute decision-making model (HMADM). The authors provided a mathematical model for optimum planning and scheduling of SH energy hubs in [297]. By employing a multiobjective evolutionary algorithm (MOEA), [298] proposed a household energy consumption planning model with the aims of minimizing the electricity costs and maximizing customer satisfaction. Framework for energy conservation system for SHs is shown in Figure 23. The reviewed literature for SHs is summarized in Table 7.

Table 7. Smart homes: reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Appliances/Load Considered	Peer-to-Peer Trading
[294,296]	SHs optimal planning (Intrusion detection)	1. (⇈) Node dis-joint paths of links 2. (⇊) Expected transmission count 3. (⇊) Planning cost 1. (⇈) Smart home product 2. (⇈) Planning performance	1. Link constraints 2. Detection quality constraint 3. Transmission constraints 4. Multipath 5. Shortest path 1. Consumer perceptions	1. WSAN model as virtual graph 2. Test SH system	1. multipath planning 2. Operational planning	1. EKMF algorithm and (GTMCF) algorithm 2. Rough knowledge-based rules and Fuzzy integral-based decision	--	--	Techno-economic	--	--
[295,297]	SHs optimal planning	1. Unfair cost distribution 2. (⇊) Household cost 3. (⇊) Installation, maintenance, and emission costs	1. DSM 2. Power balance 3. Stored energy 4. Storage cap. 5. Task duration 6. RE available 7. Reserv. time 8. Scheduling vector end time 9. Allow delay 10. Non-interruptibility 11. Utility grid max. power limit 12. Demand and supply 13. Pareto optimality	Community MG	Multiobjective optimization and Scheduling	1. Eco-trade bilinear algorithm 2. MILP	1. MG, RESs, Grid 2. WT, PV, CHP, EVs, ESS	1. Tesla Powerwall battery 2. EVs charging station, Thermal storage	Economic	1. Interruptible and non-interruptible appliances 2. Sunlight, cooler/heater, freezer, fridge, washer, dryer, dishwasher	1. Yes 2. No
[299,300,301]	Generation and load forecasting	1. Inhabitant activity prediction 2. Expected energy prediction 3. Human behavior 4. (⇈) Prediction accuracy	1. ON/OFF states 2. Inhabitant behavior 3. No. of nodes 4. Task period 5. Number of sequential events 6. Climate conditions 7. IOT dataset 8. Difference, variance and status 9. Execution time 10. Tape-and-forget data sensor	Test system	Short-term planning	1. SPEED algorithm 2. LSTM model 3. DL	--	--	Technical	1. Light, coffee maker, air conditioner, TV 2. Fridge, TV, Light, and others	--
[302,303,304,305]	Generation and load forecasting	1. Next user action forecasting 2. Appliance energy use prediction 3. Comfort level 4. (⇈) Prediction accuracy 5. (⇊) Energy consumption	1. Device characteristics 2. Action data 3. Location 4. Pattern matching 5. Environment states 6. Historical data 7. Comf. criteria 8. Economy 9. Users devises 10. Price predict.	1. SH consisting of living rooms and bedrooms 2. SH in France 3. IRISE data 4. Historical dataset of digital STROM with 33 homes	Short-term (24 h) planning	1. Pattern matching and reinforcement learning 2. Stochastic process 3. Improve learning algorithms 4. Frequent sequential pattern mining algorithm	--	--	Technical	1. 5-devices Lamp (2), Fan (1), TV(2) 2. Water heater, Freezer, Lighting Fridge 3. Lighting load 4. 3521 devices	No No No No
[306]	Generation and load forecasting	1. Consumer load prediction 2. (⇊) Users cost	1. Demand response 2. Possible schedules 3. No. of slots 4. Total power demand 5. Hour wise dispensable power	Test system with residential consumers	Scheduling	GA			Techno-economic	Residential appliances	No
[307]	Generation and load forecasting	1. SH energy prediction 2. (⇊) Energy consumption 3. (⇊) Energy cost	1. Relative humidity 2. Temperature 3. Wind speed 4. Daily energy demand	Two homes data provided by University of Massachusetts	Short-term 48- and 45-day planning	Self-organizing map (SOM)			Techno-economic	HVAC load	No
[308]	Generation and load forecasting	4. (⇈) User comfort 2. (⇊) Energy consumption	1. Available power 3. User desires 5. Temperature limits 6. Illumination limits 7. Air quality limits 8. Energy limits	Lab environment with two connected rooms	Short-term (1 month) planning	Combined GA and PSO, Kalman filter			Technical	Fan, AC, Heater, Bulb	No
[309]	Generation and load forecasting	1. GHG emission prediction 2. (⇊) GHG emissions	1. Source emission factors 2. Emission factors with different seasons and regions	Test system with multiregional data	Short-term planning	LSTM approach	Biomass, coal, gas, hydro, nuclear, oil, PV WP	--	Environment	Dishwasher EV	No
[310]	Planning and scheduling	1. (⇊) User cost 2. (⇊) PAR	1. Interruptible load 2. Non-interruptible 3. Must run load 4. Demand response	Exemplary residential system	Short-term planning	Tackling the load uncertainty method	Grid	--	Techno-economic	Five residential appliances	No
[311,312,313]	Generation and load forecasting	1. Load forecasting 2. (⇊) Forecasting error 3. (⇈) Accuracy 4. (⇈) Convergence 5. (⇊) Execution time	1. Auxiliary load 2. PV production 3. Transmission losses 4. HDD and CDD 5. Hourly electrical load 6. Temperature data	1. Texas, USA and Terni Energy, Germany 2. GEFCom2012 US utility dataset 3. 4 buildings	1. Short-term planning 2. Medium-term planning	1. PSA-DT model 2. CRBM and Jaya algo. 3. LSTM-based (seq2seq) learning model			Technical	-- 2. 5000 residential customers 3. DW, lamp, fridge radio, WM TV, laptop, PC, NAS	No No
[314]	Resource allocation	1. (⇈) Cost saving	1. Energy manag. 2. Due time of the process 3. Elect. energy consumption limit 4. Preferable resources use	Test system	Scheduling	Banker’s algorithm (BA)	PV system Grid		Economic	Shower, dishwasher, washing machine, iron, vacuum cleaner	No
[315,316]	Resource allocation	1. (⇊) User intervention 2. Peak demand management	1. Diversity factor 2. User comfort 3. Peak demand	1. Scenario based SH test system 2. Smart community with multiple SHs	Scheduling	Agent-mediated service management (AMSM)			Techno-economic	14 devices in each home e.g., TV, MW, washing, dryer, dishwash, cooker etc.	No
[317]	Resource allocation	1. Fair cost distribution 2. (⇊) Energy cost 3. (⇈) Comfort	1. Capacity constraints 2. ESS constraint 3. Energy balance 4. Starting and finishing time 5. Daily cost	Two examples of 10 and 50 SHs	Scheduling planning	MILP based lexicographic minimax method (LMM)	MG (CHP generator, TES, ESS, EVs, Grid)		Techno-economic	DW, WM, spin dryer, cooker, oven, microwave, lighting, laptop, PC, cleaner, fridge, EV	No
[318]	Resource allocation	1. RESs allocation 2. (⇊) Individual user energy cost 3. (⇊) Peak to average ratio	1. RESs 2. Min/max energy consumption 3. Contracts 4. RE production 5. Bounding and equality 6. User coupled	Interconnected users, exemplary test system	Scheduling planning	CGSD and CGF based decentralized optimization model			Techno-economic	Controllable, interruptible, and non-interruptible electrical appliances
[319]	Resource allocation	1. Bandwidth allocation 2. Fairness delay 3. Service priority	1. OpenFlow device pool 2. M2M service 3. Non M2M service 4. SDN-SH cloud and massive SH	Exemplary test system	Operational planning	Promising software defined networking (SDN) architecture			Technical	M2M and Non M2M appliances	No
[320]	Resource allocation	2. (⇊) Processing time	1. Location of data centers 2. Performance of data centers 3. Power flow 4. Data and information flow	6 geographic regions with groups of buildings with multiple SHs	Operational planning	Ant colony optimization (ACO) algorithm	MGs		Technical	IOT based devices sensor and controllers	No
[321]	Resource allocation	1. Task allocation 2. Thermal comfort	1. Virtual objects 2. Round trip time 3. Response time 4. Task dropping 5. Latency	DIY application based IOT-SH test system	Scheduling planning	Virtual objects composition (VOC) method			Technical	IOT based appliances	No
[322]	Resource allocation	1. Load balancing and stability	1. Demand response 2. PHEVs 3. Load balancing 4. Environment conditions	Typical distribution network	Scheduling planning	Support vector machine (SVM)			Technical	Residential load and EVs charging stations	No
[323]	Topology selection	1. Comfort 2. Convenience 3. Security	1. Environment conditions 2. Attention 3. Privacy 4. Priority	Test scenarios on real family of home	Scheduling planning	Human–system interaction framework			Technical	Residential controllers and sensors	No
[324]	Topology selection	1. Optimal utilization of power 2. (⇊) Power wastage	1. Event category 2. Human behavior 3. Activities 4. Data size 5. Errors	Exemplary SH system	Scheduling planning	ANN			Technical	Household appliances	No
[325,326]	Topology selection	1. Home comfort 2. Leisure 3. safety 4. Activity recognition	1. Embedded system, 3G, and ZIGBEE 2. Environment conditions	Exemplary SH	Design planning	1. IOT 2. MAS			Technical	Electrical appliances	No

Table 7. Smart homes: reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Appliances/Load Considered	Peer-to-Peer Trading
[294,296]	SHs optimal planning (Intrusion detection)	1. (⇈) Node dis-joint paths of links 2. (⇊) Expected transmission count 3. (⇊) Planning cost 1. (⇈) Smart home product 2. (⇈) Planning performance	1. Link constraints 2. Detection quality constraint 3. Transmission constraints 4. Multipath 5. Shortest path 1. Consumer perceptions	1. WSAN model as virtual graph 2. Test SH system	1. multipath planning 2. Operational planning	1. EKMF algorithm and (GTMCF) algorithm 2. Rough knowledge-based rules and Fuzzy integral-based decision	--	--	Techno-economic	--	--
[295,297]	SHs optimal planning	1. Unfair cost distribution 2. (⇊) Household cost 3. (⇊) Installation, maintenance, and emission costs	1. DSM 2. Power balance 3. Stored energy 4. Storage cap. 5. Task duration 6. RE available 7. Reserv. time 8. Scheduling vector end time 9. Allow delay 10. Non-interruptibility 11. Utility grid max. power limit 12. Demand and supply 13. Pareto optimality	Community MG	Multiobjective optimization and Scheduling	1. Eco-trade bilinear algorithm 2. MILP	1. MG, RESs, Grid 2. WT, PV, CHP, EVs, ESS	1. Tesla Powerwall battery 2. EVs charging station, Thermal storage	Economic	1. Interruptible and non-interruptible appliances 2. Sunlight, cooler/heater, freezer, fridge, washer, dryer, dishwasher	1. Yes 2. No
[299,300,301]	Generation and load forecasting	1. Inhabitant activity prediction 2. Expected energy prediction 3. Human behavior 4. (⇈) Prediction accuracy	1. ON/OFF states 2. Inhabitant behavior 3. No. of nodes 4. Task period 5. Number of sequential events 6. Climate conditions 7. IOT dataset 8. Difference, variance and status 9. Execution time 10. Tape-and-forget data sensor	Test system	Short-term planning	1. SPEED algorithm 2. LSTM model 3. DL	--	--	Technical	1. Light, coffee maker, air conditioner, TV 2. Fridge, TV, Light, and others	--
[302,303,304,305]	Generation and load forecasting	1. Next user action forecasting 2. Appliance energy use prediction 3. Comfort level 4. (⇈) Prediction accuracy 5. (⇊) Energy consumption	1. Device characteristics 2. Action data 3. Location 4. Pattern matching 5. Environment states 6. Historical data 7. Comf. criteria 8. Economy 9. Users devises 10. Price predict.	1. SH consisting of living rooms and bedrooms 2. SH in France 3. IRISE data 4. Historical dataset of digital STROM with 33 homes	Short-term (24 h) planning	1. Pattern matching and reinforcement learning 2. Stochastic process 3. Improve learning algorithms 4. Frequent sequential pattern mining algorithm	--	--	Technical	1. 5-devices Lamp (2), Fan (1), TV(2) 2. Water heater, Freezer, Lighting Fridge 3. Lighting load 4. 3521 devices	No No No No
[306]	Generation and load forecasting	1. Consumer load prediction 2. (⇊) Users cost	1. Demand response 2. Possible schedules 3. No. of slots 4. Total power demand 5. Hour wise dispensable power	Test system with residential consumers	Scheduling	GA			Techno-economic	Residential appliances	No
[307]	Generation and load forecasting	1. SH energy prediction 2. (⇊) Energy consumption 3. (⇊) Energy cost	1. Relative humidity 2. Temperature 3. Wind speed 4. Daily energy demand	Two homes data provided by University of Massachusetts	Short-term 48- and 45-day planning	Self-organizing map (SOM)			Techno-economic	HVAC load	No
[308]	Generation and load forecasting	4. (⇈) User comfort 2. (⇊) Energy consumption	1. Available power 3. User desires 5. Temperature limits 6. Illumination limits 7. Air quality limits 8. Energy limits	Lab environment with two connected rooms	Short-term (1 month) planning	Combined GA and PSO, Kalman filter			Technical	Fan, AC, Heater, Bulb	No
[309]	Generation and load forecasting	1. GHG emission prediction 2. (⇊) GHG emissions	1. Source emission factors 2. Emission factors with different seasons and regions	Test system with multiregional data	Short-term planning	LSTM approach	Biomass, coal, gas, hydro, nuclear, oil, PV WP	--	Environment	Dishwasher EV	No
[310]	Planning and scheduling	1. (⇊) User cost 2. (⇊) PAR	1. Interruptible load 2. Non-interruptible 3. Must run load 4. Demand response	Exemplary residential system	Short-term planning	Tackling the load uncertainty method	Grid	--	Techno-economic	Five residential appliances	No
[311,312,313]	Generation and load forecasting	1. Load forecasting 2. (⇊) Forecasting error 3. (⇈) Accuracy 4. (⇈) Convergence 5. (⇊) Execution time	1. Auxiliary load 2. PV production 3. Transmission losses 4. HDD and CDD 5. Hourly electrical load 6. Temperature data	1. Texas, USA and Terni Energy, Germany 2. GEFCom2012 US utility dataset 3. 4 buildings	1. Short-term planning 2. Medium-term planning	1. PSA-DT model 2. CRBM and Jaya algo. 3. LSTM-based (seq2seq) learning model			Technical	-- 2. 5000 residential customers 3. DW, lamp, fridge radio, WM TV, laptop, PC, NAS	No No
[314]	Resource allocation	1. (⇈) Cost saving	1. Energy manag. 2. Due time of the process 3. Elect. energy consumption limit 4. Preferable resources use	Test system	Scheduling	Banker’s algorithm (BA)	PV system Grid		Economic	Shower, dishwasher, washing machine, iron, vacuum cleaner	No
[315,316]	Resource allocation	1. (⇊) User intervention 2. Peak demand management	1. Diversity factor 2. User comfort 3. Peak demand	1. Scenario based SH test system 2. Smart community with multiple SHs	Scheduling	Agent-mediated service management (AMSM)			Techno-economic	14 devices in each home e.g., TV, MW, washing, dryer, dishwash, cooker etc.	No
[317]	Resource allocation	1. Fair cost distribution 2. (⇊) Energy cost 3. (⇈) Comfort	1. Capacity constraints 2. ESS constraint 3. Energy balance 4. Starting and finishing time 5. Daily cost	Two examples of 10 and 50 SHs	Scheduling planning	MILP based lexicographic minimax method (LMM)	MG (CHP generator, TES, ESS, EVs, Grid)		Techno-economic	DW, WM, spin dryer, cooker, oven, microwave, lighting, laptop, PC, cleaner, fridge, EV	No
[318]	Resource allocation	1. RESs allocation 2. (⇊) Individual user energy cost 3. (⇊) Peak to average ratio	1. RESs 2. Min/max energy consumption 3. Contracts 4. RE production 5. Bounding and equality 6. User coupled	Interconnected users, exemplary test system	Scheduling planning	CGSD and CGF based decentralized optimization model			Techno-economic	Controllable, interruptible, and non-interruptible electrical appliances
[319]	Resource allocation	1. Bandwidth allocation 2. Fairness delay 3. Service priority	1. OpenFlow device pool 2. M2M service 3. Non M2M service 4. SDN-SH cloud and massive SH	Exemplary test system	Operational planning	Promising software defined networking (SDN) architecture			Technical	M2M and Non M2M appliances	No
[320]	Resource allocation	2. (⇊) Processing time	1. Location of data centers 2. Performance of data centers 3. Power flow 4. Data and information flow	6 geographic regions with groups of buildings with multiple SHs	Operational planning	Ant colony optimization (ACO) algorithm	MGs		Technical	IOT based devices sensor and controllers	No
[321]	Resource allocation	1. Task allocation 2. Thermal comfort	1. Virtual objects 2. Round trip time 3. Response time 4. Task dropping 5. Latency	DIY application based IOT-SH test system	Scheduling planning	Virtual objects composition (VOC) method			Technical	IOT based appliances	No
[322]	Resource allocation	1. Load balancing and stability	1. Demand response 2. PHEVs 3. Load balancing 4. Environment conditions	Typical distribution network	Scheduling planning	Support vector machine (SVM)			Technical	Residential load and EVs charging stations	No
[323]	Topology selection	1. Comfort 2. Convenience 3. Security	1. Environment conditions 2. Attention 3. Privacy 4. Priority	Test scenarios on real family of home	Scheduling planning	Human–system interaction framework			Technical	Residential controllers and sensors	No
[324]	Topology selection	1. Optimal utilization of power 2. (⇊) Power wastage	1. Event category 2. Human behavior 3. Activities 4. Data size 5. Errors	Exemplary SH system	Scheduling planning	ANN			Technical	Household appliances	No
[325,326]	Topology selection	1. Home comfort 2. Leisure 3. safety 4. Activity recognition	1. Embedded system, 3G, and ZIGBEE 2. Environment conditions	Exemplary SH	Design planning	1. IOT 2. MAS			Technical	Electrical appliances	No

Figure 23. Framework for energy conservation system for SHs (adapted from [327]).

Figure 23. Framework for energy conservation system for SHs (adapted from [327]).

10.1.1. Generation and Load Forecasting

Different algorithm techniques have been reported in the literature for efficient prediction of future states, inhabitant activities, and demand prediction/forecasting of SHs. The techniques include ANN [53], sequence prediction via enhanced episode discovery (SPEED) and partial matching algorithm (PMA) [299], long short-term memory (LSTM)-based models [300], and deep learning (DL) [301]. Instead of general sequence patterns, [299] developed a human activity pattern algorithm based on psychological behavior of inhabitants for sequence and activity prediction of inhabitants. Based on the LSTM model, [300] addressed the energy consumption problem and impacts of climatic conditions by determining patterns and predicting weather characteristics. The authors in [302] proposed a pattern-matching and reinforcement learning (PM&RL)-based algorithm for predicting user action in the next state. The authors in [303] proposed a stochastic-based predictor for appliances’ energy consumption for the next day in SHs. Similarly, different approaches, such as ML-based learning algorithms [304,305], quantitative assessment-based methods [328], GA [306], an ANN based self-organizing map [307], combined GA and PSO [308], and linear regression analysis [329] approaches have been carried out in energy and user actions predictions in SHs. SHs are considered more critical for lowering domestic power usage and greenhouse gas (GHG) emissions. The authors in [309] proposed a time-series approach-based variant of recurrent neural networks (RNNs) to optimally predict GHG emissions associated with EMS of SHs. The authors in [311] utilized a scenario analysis and decision tree (PSA-DT) model for short-term forecasting of electrical energy load as demand in SHs under smart grid paradigm. The authors in [312] proposed a deep learning (DL)-based approach for predicting month-ahead hourly electrical demands by combining hourly electrical load and temperature data in SHs. An LSTM-based sequence-to-sequence learning model was developed by [313] for short-term load forecasting of SH appliances.

A comprehensive survey was carried by [330] to investigate the different types of prediction techniques for SHs.

10.1.2. Resource Allocation

Optimal resource allocation offers improvement and future recommendations in energy management in SHs. Based on Banker’s algorithm, [314] proposed a resource allocation approach by considering a set of four elements of resources: cheap electricity, expensive electricity, solar energy, and cold water in an SH framework. The authors in [331] critically discussed state-of-the-art and open research challenges in optimal planning and management of resources in SHs. Different approaches, such as agent-mediated service management (AMSM) [315], Pareto resource allocation [316], MILP-based lexicographic minimax method (LMM) [317], combined Gauss–Seidel decomposition and competitive game formulation-based decentralized optimization [318], IOT-based protocol [319], cooperative spectrum resource allocation (CSRA) framework [332], ant colony optimization (ACO) algorithms [320], virtual object composition (VOC) [321], and support vector machine (SVM) [322] models have been reported in the literature for optimal allocation of different types of resources, including cost distribution, task allocation, RES allocation, bandwidth allocation, and DR management in SHs.

10.1.3. Topology Selection

The optimal system design, as well as its management and security of electric power, plays a critical role in the economics and long-term sustainable growth of SHs. The authors in [323] developed an SH system design frame and later realized the human system interaction framework for its implementation. The authors in [324] designed an adaptive SH system for optimal utilization of electric power through ANN techniques. Using an embedded system, 3G, and ZIGBEE technologies, [325] designed an IOT-based SH system for home comfort, leisure time, and security. A unique framework for the design and deployment of SH applications focusing on activity identification in home settings was created using cloud-assisted agent-based smart home environment (CASE) architecture [326].

11. Smart Neighborhoods

From Planning Perspectives

The reviewed literature for SNHs from planning perspective is summarized in

Table 8. Smart neighborhoods: Reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources	Storage	Aspects and SDG Impact	Other Comments
[333,345]	Planning (EV Charging Station Allocation)	1. (⇈) Demand coverage 2. (⇊) Power loss 3. (⇊) Charging zone center deviation	1. Demand covered 2. No. and capacity of stations 3. Nighttime and daytime demand 4. Cap. of CS 5. No. of supply points 6. Transformer availability 7. P and Q limits 8. V and I limits	1. Neighborhood of Lisbon, Portugal, called Avenidas Novas 2. A real distribution system case	1. Long-term planning 2. multiobjective long-term planning	1. Maximal covering model 2. Hierarchical clustering method	--	Mitsubishi I MiEV Peugeot iOn Citroën C-ZERO Nissan Leaf	Technical	1. Slow charging 2. Semi-fast charging
[334]	Planning (Demand Response)	1. (⇊) Emissions 2. (⇊) Cost 3. (⇈) RESs utilization	1. Electrical and thermal equilibrium 2. Retailer constraints 3. ES and TES constraints 4. EV constraint 5. ESP, TSP, CHP, GB, DR, ATL, AEL, TD, REF, DW, HHW constraint	Neighborhood grid consisting of several H-MGs	Multiobjective planning	Metaheuristic algorithm and Taguchi orthogonal array testing (TOAT) method	MGs	Battery storage	Techno-econo-environmental	Short/long-term contract
[336]	Planning (Energy Production plants sizing and location)	1. (⇊) Installation and operation cost	1. No and capacity of generation plants 2. Economic constraint 3. Network constraints 4. Output capacities	Test case in Savona, Italy	Long-term planning	Community decision model	WT PV MT Boiler Grid	--	Economic	--
[337]	Planning (DERs)	1. (⇊) Present value of all costs (Investment, operation, maintenance, and emission costs)	1. Energy balance 2. Equipment capacity 3. Electricity selling 4. Heating pipeline network	Neighborhood consisted of 4 and 5 houses	Long-term (yearly) planning	Robust optimization approach	PV Chillers Boilers Storage tank CHPs	Heat storage	Econo-environmental
[340]	Planning (Optimal DG placement)	1. (⇈) Voltage profile 2. (⇊) Investment cost 3. (⇊) Losses	1. Load flow 2. Voltage constraints 3. Bus capacity 4. Production limits 5. P and Q limits 6. Thermal limits	IEEE 33-bus radial distribution system	Multiobjective planning	GA and PSO	WT PV	--	Techno-economic	--
[343]	Planning (Demand Response)	2. (⇊) Total electricity and gas costs	1. TOU pricing 2. EVs constraint 3. DR constraint 4. RESs constraint 5. ESS constraint 6. Temperature conditions	Exemplary neighborhood test system	Scheduling and long-term planning	Decision-making model	HPs, CHP, ESS, ACs, and RESs	Battery and thermal storage	Economic	--
[339]	Planning (Demand Side Management)	1. (⇊) Consumer cost 2. (⇊) PAR 3. (⇈) Comfort	1. RTP tariffs 2. Consumption constraints 3. LAN 4. Peak and off-peak time slots	A test case in Sydney, Australia	Scheduling planning	Game-theoretic approach and Nash-game-theory-based optimization model	PV ESS	Battery	Techno-economic	--

The authors in [333] published a study on the location of EV charging stations in a neighborhood of Lisbon, Portugal’s capital city, with a high intensity of people and employment. In [334], the authors explored the impact of a price-responsive DR program on multiobjective energy management of SNH-MG. They cut emissions and increased the revenue accomplished by home MG by 37% and 10%, respectively. The impacts of SNH load on distribution transformer in practical power systems were investigated in [335]. The authors of [336] designed and implemented a community decision model for energy production planning in Savona, Italy. Under demand unpredictability due to data scarcity, [337] proposed a robust optimization approach for development of an energy system in a neighborhood, including its pipeline network. The authors in [338] proposed a DSM approach for robust peak load shaving for a neighborhood in the presence of EVs. In [339], the authors suggested an enhanced game theory-based DSM framework for reducing end-user costs and minimizing the neighborhood’s peak to average ratio (PAR). Under the objective function of network loss reduction, [340] proposed a combined GA and PSO algorithm-based approach to address the optimal siting of wind- and PV-based DGs simultaneously with and without addressing the impacts of GHG emission. The proposed method had PARs of 1.76 and 1.81 in the summer and winter, respectively. In comparison to the non-cooperative method, the proposed model lowered costs by 9.17% in the summer and 9.68% in the winter. The authors in [341] presented a smart design technique based on the NSGA-II algorithm to help design decision making in sustainable neighborhood development with the goal of maximizing floor area ratio (FAR) and minimizing energy usage and outdoor human discomfort. A novel DR model for SNH was developed [342]. The authors in [343] developed a decision-making model to meet electrical, heating, and cooling demands through multienergy systems. The authors in [344] performed a case study for optimal planning and optimization of energy systems at neighborhood level by considering a group of 10 residential buildings in rural area of Germany. The authors in [345] offered a multiobjective Pareto frontier (PF)-based strategy to define the best location for EV semi-fast-charging stations at the neighborhood level.

12. Smart Buildings

12.1. From Planning Perspectives

Smart buildings are energy-efficient structures that provide a cost-effective way of living, a high degree of comfort, and a carbon-free environment. Building energy efficiency measures can result in considerable reductions in GHG emissions. Economic and environmental criteria should guide the design and planning of technology in SBs.

In [346], the authors assessed a planning approach for a multiobjective decision issue based on life-cycle cost assessment (LCA) estimations in order to reduce costs and GHG emissions for SBs. The researchers in [347] employed IOT technology to solve the problem of energy optimization in SBs, taking into account both the planning and operating aspects to improve user comfort. In [348], the authors formulated a MIP-based problem for optimum energy management of interconnected multi-SBs that included a CHP generator, a BESS, and a thermal storage system (TSS). The implications of power exchange capabilities were examined by examining energy flow between buildings. Based on a PSO algorithm, [349] addressed the modeling and planning of EMS in SBs with responsive/unresponsive appliances and renewable energy sources. The authors suggested a two-stage decentralized robust planning technique in [350] to study the optimum planning of interconnected SBs with bilateral transactions for optimal investment and operation. The reviewed literature from planning perspectives is summarized in

Table 9. Smart buildings: Reviewed literature from planning perspectives.

Ref.	Focus Area	Objective (s)	Major Constraints	Test System(s)	Planning Type	Applied Method	Sources/Appliances	Storage	Aspects and SDG Impact	Connected with Grid	Type of Building
[351]	Generation and load forecasting	1. (⇈) Energy consumption efficiency	1. Average daily load	Test non-residential building	Short-term planning	Post processing algorithm (PPA)	--	--	Techno-economic	Yes	Commercial building
[352]	Generation and load forecasting	1. Evaluate the perf. of ANN 2. Anal. relative performance 3. How network design parameters	1. Building type 2. Weather condit. 3. Load profile 4. Seasonal variat. 5. Power demand	Real-world data of a Campus in downtown Montreal	Short term and very short-term planning	ANN	--	--	Technical	Yes	5. Residential, Commercial and educational/office buildings
[353]	Generation and load forecasting	1. Energy consumption prediction	1. Appliance energy dataset 2. Household energy dataset	Test residential building	Scheduling planning	Hybrid DL	--	--	Technical	Yes	Residential building
[346]	Resource allocation	1. (⇊) GHG emissions 2. (⇊) Total cost	1. No. and kinds of generation units 2. Energy buy 3. PV constraints 4. WP constraints 5. MT constraints 6. Energy balance	A building condominium situated in Tortona	Long-term (20 years) multiobjective planning	LCA	PV WT MT Grid	--	Econo-environmental	Yes	Commercial building
[347,348]	Resource allocation	1. WSN deploy. 2. (⇊) Energy consumption 3. (⇈) Comfort 4. (⇊) Oper. cost	1. Connectivity 2. Protection 3. Clustering 4. Human behavior 5. Power exchange	1. Test building 2. Clustered SBs	Operational planning	1. MIP 2. MIP	-- CHP, BESS, TSS, Grid	-- Battery and Thermal	1. Techno-econo-Social 2. Economic	Yes	1. Office building 2. Apartment buildings
[349]	Resource allocation	1. (⇊) Loss 2. (⇊) Cost	1. DR 2. Buy and sell 3, Irradiance 4. Gen. limit 5. ESS constraint 6. Load profile	Modified IEEE 30-bus test system	Long-term planning	PSO	PV, ESS/chiller heater, lighting, washing machine, TV, PC	Battery	Techno-economic	Yes	Residential building
[354]	Topology selection	1. Optimal configuration and topology of UPS 2. (⇈) Energy efficiency	1. Distributed DC mode of operation 2. PQ and reliability 3. V magnitudes 4. Duration	Exemplary test system	Operational planning	PQ audits	--	Battery	Technical	Yes	High-tech building
[355]	Topology selection	1. Appliance load monitoring	1. front-end power supply circuit 2. Commer. supply 3. Prior knowledge 4. VI curve 5. Frequency 6. V-I trajectory	Exemplary system	Monitoring and scheduling	Unintrusive appliance load monitoring (NILM) strategy	Resistive load, reactive load, E-load, linear load, phase angle-controlled load	--	Technical	--	Residential building
[356]	Topology selection	2. (⇈) Thermal comfort 2. (⇈) Energy efficiency	1. Humidity 2. MRT 3. Temperature	Building in center city Philadelphia, USA	Short-term planning	Fine-grained deep learning (FGDL) approach	--	--	Techno-Social	Yes	Office building

.

12.1.1. Generation and Load Forecasting

SBs include sensors that allow them to monitor a variety of building functions, such as HVAC, lighting, and energy consumption. The data may be stored, analyzed, and further used to forecast various factors. With the recent advent of SGs, new ecosystems capable of combining demand, generation, and storage have emerged. In order to function properly, they use intelligent and adaptable features that need more advanced approaches for accurate and exact demand and generation forecasts. A thorough review was carried out [357] for debate relating to the most significant research on electric demand forecast during the previous 40 years, and showed the many models employed as well as future trends [357]. Energy consumption forecasting in SBs and utilization of obtained information to design and operate power generation are critical components of energy management in SG. Different techniques, such as a postprocessing algorithm (PPA) [351] and hybrid deep learning (HDL) ANN [352,353], have been reported for optimal short-term and long-term forecasting of power and energy in the literature. A comparative study on different energy forecasting techniques in SBs was presented in [358]. Similarly, AI-based forecasting approaches for SBs were reviewed in [359].

12.1.2. Resource Allocation

With recent advancements in the Internet of Things (IOT), IOT now offers connectivity protocols that allow new possibilities for tracking, controlling, and communicating with the environment, as well as interconnecting multiple sensors to establish a wireless sensor network (WSN), which serves as the foundation for SCs, parking lots, and emergency services, among others [360]. The author of [347] suggested a mixed integer programming (MIP)-based solution to the problem of SB energy optimization. So far, a building management framework has been designed for optimum implementation of WSN, sensors, and gateways with the aim of reducing total energy usage and increasing consumer comfort. The authors of [346] developed a decision support system (DSS) for the integration of life-cycle assessment (LCA) methodology and optimization algorithms for the optimal selection and sizing of energy-generating technologies in SBs. The authors of [347] suggested a strategy that takes into account both the planning and operational elements by optimally identifying different types of sensors and gateways for optimal deployment of a wireless sensor network (WSN) inside a building.

12.1.3. Topology Selection

The authors in [354] developed optimal configuration and topology of uninterruptible power supply (UPS) in distributed DC mode of operation to reach an energy-efficient solution in SBs. For categorization and monitoring of appliance loads in SBs, [355] focused on front-end power supply circuit topologies and the commercial power supply industry. The authors of [361] presented a method for designing and developing a topology for wellness sensor networks that is reliable, efficient, adaptable, cheap, real-time, and realistic for SH systems with the goal of expanding to SBs. Based on fine-grained deep learning (FGDL), [356] performed a data analysis of buildings to derive topology for accurate thermal comfort model for SB control. The topic of developing a topological building model for interior navigation of SBs was discussed in [362]. In [363], the authors suggested a Rubik’s cube topology (RCT) for bilevel energy transaction based on further enhancement of a two-dimensional square lattice model (TDSLM).

13. Smart Cities

13.1. From Planning Perspectives

The notion of smart cities (SCs) refers to a structure that uses a sustainable and clean energy source for its residential, commercial, transportation, and, to some extent, industrial sectors. SC energy systems need a significantly larger percentage of RESs for heat and electricity, as well as a high level of integration between industry and utilities providing consumers and businesses. The summarized literature from planning perspective is given in

Table 10. Smart cities: Reviewed literature from planning perspectives.

Ref.	Focus Area	Objective(s)	Major Constraints	Test System (s)	Planning Type	Applied Method	Energy Resources/Appliances	Storage	Aspects and SDG Impact	Connected with Grid	City Considered
[351]	Generation and load forecasting	1. (⇈) Energy consumption efficiency 1. (⇊) Loss	1. Average daily load	Test non-residential building	Short-term planning	Post processing algorithm (PPA)	--	--	Techno-economic	Yes
[336,367]	Planning (Location and size)	1. Optimal energy mix 2. (⇊) Investment and total cost	1. Capacity of units 2. No of units 3. P and Q limits	Exemplary system	Short-term planning	HOMER	PV Biogas CHP PHEV	Battery	Economic	Yes
[369]	Planning (Sizing of CSs for EVs)	1. (⇊) Overall cost 2. (⇊) Travel time 3. (⇈) Lowest PIF	1. Density traffic 2. Edge capacity constraints 3. Traffic engg. Constraints 4. Operational constraints	Test case system	Long-term planning	Multi commodity flow problem (MCFP) algorithm	Grid	Battery	Techno-economic	Yes
[370]	Generation and load forecasting	1. Price and demand forecast for big data	1. Mean absolute error 2. Normalized root means square error	NewYork City, USA	Short-term planning	DL-LSTM	Grid	--	Technical	Yes	New York, USA
[371]	Generation and load forecasting	1. Load forecasting 2. (⇊) Forecasting error	1. Actual hourly load 2. Mean absolute percentage error 3. Normalized mean-square error	6 cities	Short-term planning	LWSVR and MGOA	Grid	--	Technical	Yes	New York Victoria New England
[372]	Generation and load forecasting	1. Load forecasting	1. Heating system features of buildings 2. Seasonal consumption 3. Architectural features of buildings	Non-residential buildings in University of Girona	Short-term planning	Data driven model (DDM)	Grid	--	Technical	Yes	Girona
[373]	Generation and load forecasting	1. Forecasting Production and consumption 2. (⇊) Uncertainty	1. Solar irradiance 2. ESS constraints 3. Energy exchange 4. MAE	Test case	Short-term planning	Context-based technique	PV ESS Grid	Battery	Technical	Yes	--
[374]	Generation and load forecasting	1. Load forecasting	1. Calendar 2. Population 3. Weather data 4. Mean absolute parentage error	25 cities	Short-term planning	TL and DNN	Grid	--	Technical	Yes	Seoul
[375]	Resource allocation (Sensor Placement)	1. (⇈) Coverage 2. (⇈) Lifetime	1. Sensor and sink placement 2. Sensor and sink placement limit 3. Cell coverage 4. Cells distance 5. AND function 6. Network connectivity	A smart parking system at CERIST research center in Algiers, Algeria	Multiobjective planning	ILP	--	--	Technical	Yes	Algeria
[376]	Resource allocation	Optimize: 1. Power drain 2. Battery lifetime 3. Standard deviation 4. PLR and delay time	1. Data processing 2. Sensor networks 3. Data integration 4. Quality control	Exemplary test system	Operational planning	HABPA and DSA	--	--	Technical	Yes	Exemplary smart city
[377,378]	Resource allocation (Edge resource allocation)	1. (⇊) Average service response time	1. Latency constraints 2. Capacity constraints	Exemplary test system	Operational planning	Heuristic edge resource allocation algorithm (CHERA)	--	--	Technical	Yes	Exemplary smart city
[364]	Resource allocation (RESs allocation)	1. Optimal RESs allocation	1. Type of RESs 2. RESs capacity 3. Energy demand 4. Seasonal conditions	Case study in medium-sized city in Austria (Graz)	Long-term planning	P-Graph method	--	--	Techno-econo-environmental	Yes	Austria (Graz)

. In [364], the authors explored the application of the P-Graph methodology (PGM) to sustainable technology systems in order to build optimum energy systems that connect industry to SCs and to incorporate novel energy technologies into such systems. An extensive review on energy-related work conducted in [365] on planning and operation models within the SC with respect to five main areas includes generation, storage, infrastructure, facilities, and transport. In [366], the authors developed a technique for maximizing urban power DN utilizing idealized networks and the idealized notion. In [336,367], the authors developed a planning model for the optimal energy mix for SCs and addressed a trifold investment reduction solution for the total costs of the smart energy system installation. In [368], the authors suggested a data-driven solution based on IoT data for smart and efficient energy planning for MGs in SCs, with the goal of narrowing the gap between demand and consumption and decreasing energy losses. The transition of SHs to SCs is shown in Figure 24, while the energy conservation in SCs is summarized in Figure 25.

By considering the urban traffic flow, optimal planning for EV charging stations for SCs was studied in [369]. The roles and features of SCs are given in Figure 26.

13.1.1. Generation and Load Forecasting

DSM is the major component of an SG environment. As SGs are the center point of SCs, so DSM plays an important role in city planning, which encourages users to reschedule their energy consumption profile according to price signals received from the utility in order to reduce the cost and peak to average ratio of the system [310]. Prices and load on the demand side are strongly linked, because during peak demand, when the load profile rises suddenly, price of electricity rises and vice versa, so price forecasting is very important parameters to efficiently apply the DSM techniques at the customer end. Based on the long short-term memory (LSTM) model, [370] proposed a new price and load forecasting scheme for big data in SCs. Based on hybrid locally weighted support vector regression (LWSVR) and the modified grasshopper optimization algorithm (MGOA), [371] proposed a short load forecasting approach for SCs. A solar irradiance forecasting methodology for SCs was developed in [379].

Under the objectives of short-term load prediction, [372] identified a service required using a data-driven model (DDM) for an SC platform. The authors in [372] presented and combined a convolutional neural network (CNN) and long short-term memory (LSTM) model for optimal forecasting of air pollutants, such as particle matter ($P{M}_{2.5}$) in SCs. In [380], the authors examined spatiotemporal forecasting approaches in order to better manage critical components, such as SGs and intelligent transportation systems. The efficiency of a Markov chain (MC) model with the aim of electricity production and consumption forecasting in SCs was evaluated and tested [373]. Other approaches, such as a hybrid genetic algorithm (HGA) [371], transfer learning (TL) [374], and ML-based long short-term memory (LSTM) model were reported in the literature for short-term and monthly load forecasting in SCs. For planning reasons in SCs, most of the analysis is based on predicting time-series data collected by smart sensors. While forecasting in SCs has its advantages, it also has its own set of challenges, such as multistage forecasting challenges, low-quality training data, malfunctioning, or communication letdowns, and predictive value in systems with increasing numbers of sensors [381].

13.1.2. Resource Allocation

One of the most prominent characteristics of SCs is optimal resource allocation. As a field of operations research and computational geometry, location analysis is concerned with finding resource placements while optimizing various objectives, such as cost, waiting time, trip distance, or coverage [382]. The authors in [375] proposed an integer linear programming-based multiobjective model for optimal sensor placement for the design of smart parking networks in SCs: Mobi-Het, or mobility-aware optimum resource allocation architecture for distant big-data task execution in mobile form

Cloud computing (MCC) was presented in [383] to achieve greater efficiency in timeliness and dependability. IoT is a new technology for SCs that links numerous digital devices over the internet, allowing for a variety of innovative services ranging from the academics to the industrial sector. The study in [376] presented the hybrid adaptive bandwidth and power algorithm (HABPA) and the delay-tolerant streaming algorithm (DSA) for optimal resource planning for IOT-based SCs. With an objective of average service response time (ASRT) minimization, [377] studied optimal allocation of edge resources in IOT-based SCs. A modified PSO framework was developed with the aim of maximizing covering location by optimally placing the demand in the network [382]. Other approaches, such as the modified fish swarm algorithm (MFSA) [384], VM and Nash equilibrium [378], and maximum reward algorithm (MRA) [385] have been reported in the literature for optimal allocation of resources in the SC environment. Energy systems for SCs would necessitate a considerably larger proportion of RESs for heat and electricity, as well as a high level of integration of industry and utilities providing families and businesses. Based on a P-graph method, [364] proposed optimal RE system design for creating a linkage between industry and SCs.

14. Policies and On-Ground Projects for Smart Energy Ecosystem Realization

Under the SG paradigm, SDMs offer machine-to-machine (M2M) communication infrastructures that have attracted much attention in recent years due to their potential to change the complete energy ecosystem. Many developed and developing countries across the globe have formulated policies on different aspects of SG technologies. The review of different ongoing and developed policies of different nations are of prime importance for implementation of SDM concepts. Practical and on-ground projects of different SDMs across the globe are summarized.

Depending upon the geographical location, geopolitical trends, and financial concerns, the policies of different countries towards SDM realization vary with respect to different levels. SDM realization policies of any country tend to focus on the UN Sustainable Development Goals (SDGs) and to transform the traditional electricity market and network infrastructure to a sustainable and reliable one. An overview of different policies of nations is given in

Table 11. Reviewed policies of developed and developing countries across globe in terms of smart grid deployment.

Countries	Policy and Main Guidelines	Policy Aims and Objectives
Pakistan	Alternative Energy Development Board (AEDB) - Alternative Energy Policy (2019) [386] - Alternative Energy Development Board (Certifications) Regulations (2021) [387] China Pakistan Economic Corridor (CPEC) - Power grid transformation - Technical Policy Framework for Grid Transformation of CPEC Ambarella [388]	- 20% Renewable Energy Injection by (2025) - 30% Renewable Energy Injection by (2030) - Secured and quality-assured delivery of solar and wind energy projects - Renewable share enhancement - Smart grid implementation plan for 2030 - Grid modernization through different SG technologies deployment
India	Smart Grid Vision for India - Puducherry Smart Grid Pilot Project [389]	- Access, availability, and affordability of quality power for all - Secure, adaptive, sustainable, and digitally enables energy ecosystem
European Union	SG legislation and regulations in EU - Conclusions of the European Council of February 4, 2011 [390] - Commission Recommendation on Preparations for the Roll-out of Smart Metering Systems (C/2012/1342) [391] - EC Standardization Mandate for SMs (M/441) [392] - EC Standardization Mandate for EVs (M/468) [392] - EC Standardization Mandate for SGs (M/490) [392] Vision 2030 - October 2014	- Sustainable, competitive, and secure energy supply - Smart meters deployment - EVs deployment - Demand response and dynamic pricing - Standards development - Renewable penetration enhancement - Energy efficiency enhancement up to 27% and 40% reduction in GHG emissions - Ambition 2050 target to reduce GHG emissions by 95%
United States	Smart grid legislative and regulatory policies - Energy Independence and Security Act of 2007 [393] - Grid 2030 [394] - GridWise Program of US DOE [395] - NIST IOP Framework Roadmap 4.0 [396] Energy provisions in the CARES Act [397]	- Power quality and reliability - Efficiency - Sustainable development goals (SDGs) - GHG emissions reduction - Security and optimization
United Kingdom	Department of Energy and Climate Change - Smart Grid Vision and Routemap [398] - A legally binding commitment (2020 and 2050) [399] - Energy Bill 2010–11 [400] - Electricity market reform - Energy White Paper 2011 - Renewable energy guarantees of origin - Renewables obligation - Climate change levy	- All electricity network stakeholders’ involvement - Challenges facing electricity distribution networks - Demand elasticity - 30% carbon emission reduction by 2020 - 80% carbon emission reduction by 2050 - Providing incentives to consumers
South Korea	Smart Grid Roadmap - 2010–2030 [400] - Special Act on the Establishment and Support for Smart Power Grid [401] Korean New Deal Clean Energy Spending [397]	- Smart power grid - Smart consumer - Smart transportation - Smart renewable and smart electricity services - Blackout time reduction
China	12th Five-Year Plan for Energy Science and Technology Development [402] - Environment - Technology improvement - Power supply efficiency - Renewable energy resources State Grid Corporation of China 14th Five-Year Plan [397]	- Renewable energy reductions in taxes - GHG emissions reduction - Transmission and distribution - Renewable penetration enhancement - Reduce CO2 intensity of economy by 18% from 2021 to 2025 - Reduce energy intensity of economy by 13.5% from 2021 to 2025 - 20% non-fossil share of energy mix by 2025 - 25% non-fossil share of energy mix by 2030
Japan	6th Strategic Energy Plan 2021 [403] - 2021–2030 Roadmap to Intl. Std. SG SG policy by Ministry of Economy, Trade, and Industry Public spending on clean energy innovation—2021 national budget [397]	- Reduce the GHG emissions by 46% up to 2030 - Double the utilization of renewable energy - Focus on solar and wind rather than biomass
Italy	National energy strategy Net metering RES promotion—decree implementing Directive 2001/77/EC Reorganization of energy sector regulation ENEL voluntary agreement	- Energy cost minimization - Renewable penetration enhancement - Demand side management - GHG emission reduction
Australia	Australian smart grid policy and standards [404] Vision 2020 Energy policy frame Work by council of Australian Government DER-friendly policy	- Renewable energy penetration - TOU tariffs infrastructure - Security and Power Quality - Power delivery efficiency - Demand side management
Canada	National SG Technology and Standards Task Force Smart Grid Canada Energy provisions in the 2020 Healthy Environment and a Healthy Economy Plan [397] Spending in the Hydrogen Strategy and Strategic Innovation Fund Net Zero Accelerator [397]	- Renewable energy penetration enhancement - GHG emission reduction - Security and power quality
Germany	Germany’s climate policy [405] Climate Action Programme 2030 [405]	- Renewable penetration enhancement - GHG emission reduction - Energy consumption reduction
Saudi Arabia	National Strategy for Smart Meters and SGs [406] National Renewable Energy Program 2019 Energy and Sustainability (Vision 2030) [407]	- Determining the challenges facing the electricity industry - Smart meters and smart grids technologies - Reliability, quality of service, and efficiency - Better utilization of assets - RESs integration enhancement - Electricity consumption conservation - Install 58.7 GW renewable by 2030
South Africa	Smart Grid 2030 Vision [408]	- 20% sustainable reduction in peak energy demand - 100% grid availability - 40% improvement in system efficiency - 8 GW Renewable energy integration - Improved service delivery and service reliability
International Collaboration	IRENA REmap 2030 [409] 2030 National Environmental Policy [410] IEA NetZero by 2050 [411] IEA Smart Grids Technology Roadmap 2050 [412] IEA Vision 2035 [413] IEEE Smart Grid Vision for Vehicular Technology 2030 [414] IEEE Grid Vision 2050 [415] US DOE Wind Energy Vision 2030 [416] EnergyVision 2030 (Grid Modernization) [417]	- Doubling the share of renewable energy share in world final energy consumption - Energy efficiency - GHG emission reduction - Renewable penetration enhancement - Smart grid technologies deployment - Increase electrification rate - EVs trends enhancement (sustainable transportation) - Future energy availability, production, and consumption - 20% wind energy integration into the system up to 2030 - Demand optimization, Demand response, Active load management, Storage

The summary of different on-ground projects is given in

Table 12. On-ground pilot projects of smart distribution networks.

SDM	Pilot Project Name	Organization	Country	Year	Anticipated Objectives/Motives	Elements
ARDNs [418]	22 kV radial distribution network—ADN modeling for enhancing PS performance.	Egyptian Electricity Holding Company, Middle Egypt DisCo.	Egypt	2020	Sustainable power system performance
LDNs [419,420]	1. 3-terminal SOP-based pilot project of MV flexible DN as CLDN. 2. Belgium east loop active network management.	1. Beijing Electric Power Corporation, China 2. Ores and Elia	1. Yanqing district, Beijing, China 2. Belgium	1. 2018 2. 2010–11	1. Penetration accommodation of renewable generations such as solar, biomass, geothermal, wind, and hydro. 2. Load management, power quality and stability, and switching operation.
MDNs [144,421,422,423]	1. LV dense-mesh municipal distribution network 2. GridBox pilot project 3. ESB, smart green circuits, networks-SG demonstration project 4. Active network management with 33 kV distribution network.	1. E.ON Distribuce 2. Supercomputing Systems AG, Zurich 3. ESB Networks 4. Scottish Power energy networks in collaboration with Psymetrix	1. Czech Republic 2. Switzerland 3. Ireland 4. Isle of Anglesey, North Wales	1. 2019 2. 2016 3. 2010–12 4. 2013	1. Safety and durability, operational analysis, power quality monitoring 2. Monitoring and active control of DN 3. Asset utilization, power supply, system losses, system voltage, reliability, switching operation 4. Distribution network capacity, renewable penetration
MGs [424,425,426,427,428,429]	1. University of Cyprus (UCY) Nanogrid Solution 2. Smart Polygeneration Microgrid (SPM) test-bed facility at the Savona Campus of the Genoa University 3. Demonstration projects by Japan - Aichi Microgrid - Hachinohe Microgrid - Kyotango Microgrid - Sendai Power Quality Management - Shimizu Construction Company - Miyako Island Microgrid - Higashida Co-generation - New Mexico—Los Alamos - New Mexico—Albuquerque 4. European FP 5 Microgrids program—a pilot installation on Kythnos Island, Greece. 5. More Microgrids Project—a pilot installation in Mannheim-Wallstadt, Germany 6. North American microgrids - AEPCERTS - Mad River - BC Hydro Boston Bar - GE Microgrid	1. University of Cyprus 2. University of Genova 3. - NEDO - NEDO - NEDO - NEDO - Private own - Utility - Steel company - Distribution utility and NEDO - Building owner and NEDO 4. European Union MGs funding program 5. National Technical University of Athens (NTUA) 6. - Sandia National Laboratories and the University of Wisconsin - Northern Power Systems - British Columbia Hydro - General Electric	1. Cyprus 2. Italy 3. Japan 4. Greece 5. Mannheim–Wallstadt, Germany 6. US	1. 2020 2. 2013 3. - 2003–7 - 2003–7 - 2003–7 - 2004–7 - 2006 - 2009–13 - 2010 - 2010–13 - 2010–13 4. 2001–6 5. 2006 6.	1. Model commercial and residential load Techno-economic analysis 2. Focus on power and communication infrastructure 3. - Balancing - Balancing and power quality - Balancing - Balancing - Balancing and power quality - Balancing and power quality - Balancing - Ancillary services and balancing 4. Test centralized and decentralized control strategies for islanding 5. Study alternative methods, strategies along with universalization and plug-and-play concepts 6. PV distributed among homes	3. - PV, NAS battery, FC, SM, PMU. - Wind, PV, LA battery, gas engine, SM, PMU. - Wind, PV, LA battery, methane fermentation, fuel cell, SM - PV, LA battery, capacitor, city gas, FC, SM. - PV, Ni-MH battery, city gas, SM - Wind, PV, NAS battery, SCiBT, gas turbine and thermal - Wind, PV, Li-ion battery, FC, EV, SM - PV, NAS battery, LA battery, SM - PV, LA battery, city gas, FC, SM 4. Solar, gas/diesel, storage 5. PV, DG, controllable loads 6. CHP, MG, ESS - Biodiesel, MT, propane
MGs [429,430,431]	7. Hsinchiang Microgrid 8. MVV Energie Projects 9. KERI microgrid 10. DER-IREC 22@Microgrid	7. Mitsubishi Electric 8. MVV 9. Korean Energy Research Institute (KERI) 10. GTD Sistemas de Informacion SA	7. China 8. Germany 9. Korea 10. Spain	-- -- -- 2009–11	7. Serve the peak load of 90 kW. 8. To serve the residential and commercial units of load. 9. Testing and studying 10. New research platform, new components’ integration, DER and EV integration	7. Battery, PV, generator, utility 8. Utility, FC, PV, MG, flywheel, CHP 9. PV, FC, DG, WT, Storage 10. WT, PV, hydrogen, storage, utility
MMGs [44,432,433,434]	1. IIT (Illinois Institute of Technology) DC MMGs 2. NREL AC-DC hybrid MMGs 3. Kythnos single-phase/three-phase MMGs 4. Zhuhai Wanshan Islands MMGs 5. Guangxi Maoershan MMGs	1. Illinois Institute of Technology (IIT) 2. National Renewable Energy Laboratory 3. European Union Project (more microgrids) 4. -- 5. --	1. US 2. US 3. Greece 4. China 5. China	1. 2014 2. 2010 3. 2008 4. 2018 5. 2018	1. Islanding, load sharing, and resynchronization in island mode. Load sharing between north and south substation. 2. Hybrid power integration 3. Power system reliability research and upper scheduling management and ILM. 4. MMGs hierarchical control and mode switching study. 5. Research on microgrid operating mode, form of technology and engineering design.	1. WT, PV, battery, gas turbine synchronous generator, utility grid. 2. PV, conventional gen., BESS, AC load. 3. PV, BESS, DG. 4. Cage induction gen., PV, COG, LAB. 5. WT, PV, COG, and SC-ESS.
VPPs [435,436,437,438,439,440]	1. Reflexe pilot project 2. Fenix 3. Pilot VPP implementation into a vertically integrated utility system 4. GAD 5. Smart power system—first trial 6. Virtual power plant	-- 2. Iberdrola Distribution 3. Dubai national vertically integrated utility 4. Iberdrola Distribution 5. Energy Research Center of The Netherlands 6. RWE DAG DE	-- 2. Spain 3. Dubai 4. Spain 5. Netherlands 6. Germany	-- 2. 2005-9 -- 4. 2005-9 5. 2007-7 6. 2008-10	1. Forecast of expected load, production, and flexibility from participating DERs. status of power supply, demand, and flexibility available in the system 2. Large-scale VPP decentralized management, development of communication, normal/abnormal operations, integration with management and market, DG and DER penetration. 3. Practically test for communication errors, validate TVPP functionalities, demonstrate currently available control modes, effective integration of VPP into grid, operational modes, and integration scenarios. 4. Optimize energy consumption, minimize associated costs, focus on DSM, maintain quality standards, home application. 5. Reduce local peak load and improve efficiency of system. 6. Economic and technical feasibility of VPP, decentralized power production with DGs.	-- 2. Distributed energy resources, CHP, PV. 3. PV and BESS, utility-scale BES, MG, and EV charging stations 4. Distributed energy resources available in the region and storage system. 5. Micro CHP, PV, BESS, WT. EVs. 6. CHP, WT, biomass
SHs [441,442]	Energy@home	Indesit, Enel Distribuzione, Telecom Italia, Electrolux	Italy	2009-11	Energy cost incentives, informs consumer with mobile device, smart appliances, adjust demand patterns in favor of consumer, home application for consumer behavior.	ICT and AMI, advanced distribution automation, sensing, measurement, and monitoring, DR, and DSM technologies
SNHs [443]	1. Sidewalk Toronto Project 2. Lower Manhattan smart neighborhood pilot	1. Waterfront Toronto and Sidewalk Labs 2.–	1. Toronto 2.–	1. 2017–18 2.–	1. SNH planning, planners’ roles with respect to smart and digital technology change. 2.–	1. SCs, urban planning, urban planner roles, smart and digital technologies. 2.–
SBs [438,441,444]	1. Thames Valley Vision Project 2. China’s first automated demand response project 3. BeyWatch	-- 2. US Trade and Development Agency, China Electric Power Research Institute, Tianjin Economic Technological Development Area 3. Investigacion y Desarrollo SA	1. Town of Bracknell and surrounding area, UK 2. China 3. Spain, UK, Slovenia, Italy, Greece	1. 2014 2. 2011 3. 2008–11	1. Ensuring a high-quality and affordable future electricity network. 2. Energy consumption reduction, cost reduction and network reliability. 3. Develop user-centric and energy aware solution, monitor, control, and balance demand, consumer-aware energy consumption, enabling intelligent control.	1. ADR techniques, building management systems (BMSs) 2. DR, ADR 3. ICT (AMI), advanced distribution automation, sensing, measurement, and monitoring, DR/DSM.
SCs [439,445,446,447,448]	1. Smart city pilot projects developed in Amsterdam 2. Model City Manheim 3. Yokohama Smart City Project 4. Colorado Smart City Project	1. Amsterdam Smart City platform 2. MW Energie (DE) 3.– 4. Boulder, CO	1. Netherlands 2. Germany 3.– 4. US	1. 2014–16 2. 2008–12 3.– 4.–	1. Address urban sustainability issue, improve the effectiveness of urban services, and enhance the quality of life of citizens. 2. Renewable penetration and decentralized sources, project deployment, show, translate and applied to other regions. 3. Renewable integration, GHG emission reduction, PS stability, smart transportation, smart homes, smart buildings. 4. Exploitation of SG tools, implementation of DSM, consumer participation.	1. Roll out (smart meter deployment), expansion (add functionalities), replication (tested V2G system) 2. ICT, advanced distribution automation, distributed generation. 3. EVs, ICT, advanced distribution automation 4. ICT, DR/DSM.

.

15. Research Gaps, Challenges, and Recommendations

The challenges and issues we face in the current era in adoption of SG technologies, different aspects of SDMs at a practical level, and recommendations to cope these challenges are arranged as follows from Section 15.1 to Section 15.4. Moreover, the research gaps present in the existing literature are also summarized briefly with respect to different optimization techniques and other planning approaches.

15.1. Challenges in Adoption of Demand-Side Management (DSM)

DSM is a vital component of SG technologies, allowing end users to optimize their energy consumption with mutual aims and utility in terms of reducing energy cost and PAR. In developing nations, 80–90% of the features of smart meters are still not enabled due to deficiency of sufficient communication bandwidth. For example, in Pakistan, smart meters are deployed by the Islamabad Electricity Supply Company (IESCO) in some areas of Islamabad. However, not all functions of smart meters are active right now, and there are various challenges in adoption of DSM due to lack of sufficient bandwidth and public awareness. In addition, implementation of various DSM technologies needs a local battery storage system at the customer end, which can be a major barrier for DSM adoption due to cost concerns for both utility and end-user stakeholders. Other challenges or issues in deployment of DSM techniques include insufficient deployment of smart meters, lack of insufficient policies on DSM, tariff structure, load growth and balancing, less consumer awareness, incentive infrastructure, cost and finance issues, response fatigue, investment cost issues, and promotional issues at the utility end [449].

In terms of open research challenges and gaps, there is still room to improve optimization techniques. In industrial demand-side management, several key challenges are identified in the literature, i.e., accurate modeling of operational flexibilities, integration of production and energy management, long-term optimization across different time horizons, and decision making under uncertain production and pricing levels in the market [450].

15.2. Challenges in Resource Allocation

Much research has been carried out already on resource allocation problems in SDMs. A lot of studies have investigated optimal siting, sizing problem of DGs and other compensation, switching, and protection devices at the distribution end. Now, the major focus of the scientific community is on storage, which can act as load during off-peak hours (when prices or low) and behave as a generating source during peak hours (when prices are high). Though a good amount of research is present in the literature addressing storage issues, there is a still huge gap. The major focus is on battery storage technologies; however, hydrogen storage is an emerging field and can be more sustainable and cleaner in terms of energy supply. The energy can be stored in hydrogen form for a long time (for many years), and can then be converted back to electricity from hydrogen. The major challenge in terms of a hydrogen economy is power density: it has excellent energy density, but not good power density compared to battery storage.

As such, there is a large research gap present for future researchers to enhance the power density of hydrogen-based storage systems.

15.3. Forecasting Challenges

Most of the current and reviewed literature offers short-term forecasting. Most of the ANN- and DL-based models are based on short-term generation and load forecasting models of SDMs. This short-term forecasting does not facilitate the real deployment of DSM techniques, because DSM techniques are preferred over long-term planning (around 3 years or above). Moreover, short-term pricing, generation and load forecasting do not facilitate proper engagement for long-term energy purchasing contracts between stakeholders. In addition, short-term forecasting does not allow transmission, distribution, and other financial planning.

15.4. Challenges in Optimization Techniques

Various optimization and scheduling techniques have been developed and deployed by different research groups in the literature. However, almost every approach has some demerits in terms of accuracy, complexity, convergence rate, or evaluation of local or global optima. Different optimization and scheduling techniques with their challenges and issues utilized in different aspects of planning of SDMs are summarized in

Table 13. Different optimization and scheduling techniques, planning applications, and challenges or issues.

Categories	Techniques and Algorithms	Applications in Planning	Challenges
Traditional Approaches	1. Nonlinear programming (NLP) 2. Linear programming (LP) 3. Mixed integer programming (MIP) 4. Mixed integer linear programming (MILP) 5. Mixed integer nonlinear programming (MINLP) 6. Quadratic programming (QP) 7. Network flow programming (NFP) 8. Generalized reduced gradient (GRG) 9. Newton method (NM) 10. Interior point methods, etc.	1. Reactive and active power planning 2. Generation expansion planning 3. Generation and distribution planning 4. Generation and load scheduling 5. Maintenance scheduling 6. DG and other RES generation mix planning 7. Compensation placement and sizing 8. Voltage regulation 9. Long-term and short-term generation and load forecasting	1. High processing time required 2. Large number of iterations 3. Lowe convergence rate 4. Low accuracy 4. Suitable for only small problems 5. Risk for high dimensionality 6. Low flexible with varying time periods 7. Not more suitable for inequalities
Heuristic and Metaheuristic Techniques	1. Simulated annealing method (SA) 2. Tabu search algorithm (TSA) 3. Genetic algorithm (GA) 4. Evolutionary algorithm (EA) 5. Memetic algorithms (MA) 6. Particle swarm optimization (PSO) 7. Ant colony algorithm (ACA) 8. Differential evolution (DA) 9. Harmony search algorithm (HSA) 10. Honey-bee colony optimization (HBCO), etc.		1. Lower accuracy 2. Numerical inefficiency 3. Local optima 4. Low convergence rate 5. High number of iterations 6. Lot of tunable parameters 7. High complexity and time-consuming 8. Slow processing
AI-Based Techniques	1. Artificial neural networks (ANN) 2. Machine learning and reinforcement learning (ML&RL) 3. Deep learning (DL) 4. Genetic algorithm (GA) 5. Particle swarm optimization (PSO) 6. Multiagent system (MAS) 7. Fuzzy logic (FL) 8. Pareto multiobjective approach (PMA) etc.		1. Large data set required 2. Unknown network duration 3. Overloading 4. Facing difficulty at industrial deployment 3. High time requirements 5. Poor computation 6. Accuracy issues 7. Total dependent on human knowledge and expertise
Hybrid AI Techniques	1. Fuzzy neural network systems (FNNs) 2. Fuzzy neural/expert/genetic systems 3. Simulated annealing (SA) with fuzzy/genetic/expert systems etc.		3. Repeated annealing 4. Regularly updates required
Analytical Techniques	1. Salp swarm optimization (SSO) 2. Whale optimization algorithm (WOA) 3. Sensitivity analysis (SA) etc.		1. Computation complexity 2. Deal on deterministic variables 3. Does not consider probabilistic parameters 4. Slow convergence speed

.

16. Bibliometric Analysis

A bibliometric analysis of all Scopus-indexed articles was conducted through VOSviewer software. The bibliometric survey is detailed in Table 13,

Table 14. Top 32 countries with respect to papers published and citations received.

Sr. No.	Country	Documents	Percentage Document	Citations	Citation per Document	Total Link Strength	Nominal GDP Rank
1	United States	53	11.09	5910	111.51	100	1
2	China	51	10.67	1680	32.94	66	2
3	India	46	9.62	1330	28.91	45	6
4	Iran	38	7.95	1590	41.84	60	22
5	South Korea	24	5.02	978	40.75	86	10
6	Italy	23	4.81	3121	135.7	21	8
7	United Kingdom	22	4.6	2394	108.82	17	5
8	Spain	19	3.97	1725	90.79	37	14
9	Australia	17	3.56	926	54.47	17	12
10	Pakistan	16	3.35	764	47.75	72	46
11	Canada	14	2.93	3347	239.07	42	9
12	Egypt	13	2.72	216	16.62	8	36
13	Saudi Arabia	13	2.72	737	56.69	35	19
14	Portugal	12	2.51	459	38.25	11	49
15	Malaysia	11	2.3	1527	138.82	25	38
16	Netherlands	10	2.09	657	65.7	7	17
17	Turkey	10	2.09	427	42.7	14	20
18	France	9	1.88	511	56.78	12	7
19	Greece	9	1.88	1083	120.33	18	52
20	Romania	8	1.67	529	66.13	7	47
21	Brazil	7	1.46	148	21.14	0	13
22	Germany	7	1.46	131	18.71	3	4
23	Japan	7	1.46	387	55.29	6	3
24	Taiwan	7	1.46	115	16.43	10	21
25	Denmark	5	1.05	248	49.6	11	37
26	Finland	5	1.05	148	29.6	8	44
27	Switzerland	5	1.05	1057	211.4	2	18
28	South Africa	4	0.84	242	60.5	2	42
29	Sweden	4	0.84	289	72.25	3	24
30	Indonesia	3	0.63	49	16.33	0	16
31	Poland	3	0.63	24	8	5	23
32	Singapore	3	0.63	33	11	0	39

,

Table 15. Papers published and citations received by top 25 institutions.

Sr. No.	Organization	Documents	Citations	Avg. Citation per Document	Total Link Strength
1	Department of Electrical and Computer Engineering, College of Information and Communication Engineering (CICE), Sungkyunkwan University (SKKU), Suwon, South Korea	6	61	10.17	22
2	Department of Electrical and Computer Engineering, University of Denver, Denver, United States	4	409	102.25	4
3	Department of Computer Science, COMSATS University Islamabad, Islamabad, 44000, Pakistan	3	64	21.33	0
4	Department of Electrical Engineering, Azarbaijan Shahid Madani University, Tabriz, Iran	3	90	30	3
5	Department of Electrical Engineering, National Institute of Technology, Kurukshetra, India	3	34	11.33	2
6	US–Pakistan center for Advanced Studies in Energy (USPCAS-E), National University of Sciences and Technology (NUST), Islamabad, Pakistan	3	45	15	18
7	Department of Electrical and Computer Engineering, Auburn University, Auburn, United States	2	156	78	2
8	Department of Electrical and Computer Engineering, University of Alberta, Edmonton, Canada	2	367	183.5	3
9	Department of Electrical Engineering, Amirkabir University of Technology, Tehran, Iran	2	47	23.5	1
10	Department of Electrical Engineering, Dr. B.C. Roy Engineering College, Durgapur, West Bengal, India	2	124	62	3
11	Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands	2	29	14.5	0
12	Department of Electrical Engineering, Tsinghua University, Beijing, China	2	6	3	0
13	Energy Systems Division, Argonne National Laboratory, Lemont, United States	2	134	67	1
14	Faculty of Electrical and Computer Engineering, Tarbiat Modares University, Tehran, Iran	2	84	42	0
15	Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran	2	98	49	0
16	Galvin Center for Electricity Innovation, Illinois Institute of Technology, Chicago, United States	2	371	185.5	5
17	IEEE	2	81	40.5	0
18	National Technical University of Athens, Greece	2	432	216	0
19	School of Electrical and Computer Engineering, National Technical University of Athens (NTUA), Athens, Greece	2	154	77	8
20	School of Electrical Engineering, Computing, and Mathematical Sciences, Curtin University, Perth, Australia	2	18	9	0
21	School of Electrical Engineering, Southeast University, Nanjing, China	2	57	28.5	0
22	School of Technology and Innovations, University of Vaasa, Vaasa, Finland	2	40	20	1
23	State Key Laboratory of Power Systems, Department of Electrical Engineering, Tsinghua University, Beijing, China	2	36	18	3
24	UNIDEMI, Department of Mechanical and Industrial Engineering, Faculty of Science and Technology (FCT), Universidade Nova de Lisboa, Caparica, Portugal	2	18	9	0
25	University of Cagliari, Cagliari, Italy	2	6	3	0

,

Table 16. Documents published and citations received by top 43 authors.

Sr. No.	Author	Documents	Percentage Document	Citations	Citation per Document	Total Link Strength
1	Kazmi S.A.A.	11	6.83	82	7.45	59
2	Shin D.R.	8	4.97	72	9	59
3	Kumar A.	6	3.73	73	12.17	15
4	Shahidehpour M.	6	3.73	681	113.5	38
5	Khodaei A.	5	3.11	624	124.8	7
6	Li J.	5	3.11	43	8.6	4
7	Zhang X.	5	3.11	212	42.4	23
8	Carro B.	4	2.48	636	159	21
9	Chen Y.	4	2.48	165	41.25	4
10	Hatziargyriou N.	4	2.48	710	177.5	0
11	Liu Y.	4	2.48	51	12.75	4
12	Lloret J.	4	2.48	636	159	21
13	Wang J.	4	2.48	193	48.25	8
14	Wang Y.	4	2.48	172	43	8
15	Zhang Y.	4	2.48	130	32.5	6
16	Aguiar J.M.	3	1.86	494	164.67	17
17	Ahmad H.W.	3	1.86	6	2	29
18	Arefifar S.A.	3	1.86	486	162	26
19	Catal? J.P.S.	3	1.86	65	21.67	0
20	Che L.	3	1.86	412	137.33	24
21	Du Y.	3	1.86	79	26.33	1
22	Dumbrava V.	3	1.86	16	5.33	0
23	El-Fouly T.H.M.	3	1.86	378	126	20
24	Ganguly S.	3	1.86	92	30.67	0
25	Godina R.	3	1.86	43	14.33	0
26	Gupta A.R.	3	1.86	25	8.33	11
27	Haghifam M.R.	3	1.86	190	63.33	2
28	Hernandez L.	3	1.86	497	165.67	19
29	Javaid N.	3	1.86	64	21.33	0
30	Liu Z.	3	1.86	131	43.67	13
31	Mohamed Y.A.R.I.	3	1.86	486	162	26
32	Murty V.V.S.N.	3	1.86	46	15.33	6
33	Muyeen S.M.	3	1.86	28	9.33	1
34	Nojavan S.	3	1.86	192	64	1
35	Ravi K.	3	1.86	117	39	12
36	Salehi J.	3	1.86	196	65.33	6
37	Wang X.	3	1.86	424	141.33	6
38	Wang Z.	3	1.86	89	29.67	2
39	Wu W.	3	1.86	38	12.67	7
40	Xu J.	3	1.86	65	21.67	6
41	Yang J.	3	1.86	85	28.33	9
42	Yang Y.	3	1.86	67	22.33	1
43	Yuvaraj T.	3	1.86	117	39	12

and

Table 17. Work published by top 25 sources, cited in other sources.

Sr. No.	Source	Documents	Citations	Avg. Citation per Document	Total Link Strength
1	Energies	23	378	16.43	43
2	IEEE Transactions on Smart Grid	17	1945	114.41	28
3	International Journal of Electrical Power and Energy Systems	17	1117	65.71	20
4	Energy	13	1667	128.23	10
5	Renewable and Sustainable Energy Reviews	11	1883	171.18	13
6	IEEE Transactions on Power Systems	10	904	90.4	16
7	Sustainability (Switzerland)	10	120	12	9
8	IEEE Access	9	122	13.56	15
9	IET Generation, Transmission and Distribution	9	345	38.33	16
10	Applied Energy	7	387	55.29	6
11	Electric Power Systems Research	7	331	47.29	9
12	Energy and Buildings	7	343	49	7
13	Applied Sciences (Switzerland)	6	53	8.83	4
14	International Transactions on Electrical Energy Systems	5	47	9.4	9
15	Electric Power Components and Systems	4	45	11.25	3
16	Energy Procedia	4	84	21	0
17	IEEE Power and Energy Society General Meeting	4	76	19	0
18	IEEE Transactions on Industrial Informatics	4	186	46.5	2
19	IET Renewable Power Generation	4	200	50	4
20	Sustainable Cities and Society	4	544	136	4
21	Electrical Engineering	3	21	7	1
22	Energy Conversion and Management	3	301	100.33	4
23	IEEE-PES Innovative Smart Grid Technologies Conference Europe	3	16	5.33	0
24	Journal of the Institution of Engineers (India): Series B	3	21	7	1
25	Journal of Urban Technology	3	2766	922	2
26	Sustainable Energy, Grids and Networks	3	60	20	4

and Figure 27, Figure 28, Figure 29, Figure 30, Figure 31, Figure 32, Figure 33, Figure 34, Figure 35, Figure 36 and Figure 37. Table 13 showing the total documents and citations per document received by the 32 most active countries. Total ink strength shows the collaboration of countries with other countries. The US has received the most citations overall, while Canada received the most per document, which shows the quality of research published by Canada. Table 14 shows that the Department of Electrical and Computer Engineering, University of Alberta, Edmonton, Canada published few papers, but they received the highest average of citations, which shows the quality of research at that institute.

Table 15 shows that N. Hatziargyriou published few documents, but received the most citations/document, which shows his/her quality of research. Table 16 shows that IEEE Transactions on Smart Grid and Renewable and Sustainable Energy Reviews had better performance in quality research publications.

Table 17 shows that [40] received the most citations but had no connection with other documents or sources, while document [60] had the highest collaboration with other sources. The survey of top 34 documents that have received highest citations summarized in

Table 18. Survey of top 34 documents who have received more citations (citation/doc = 20).

Sr. No.	Document	Citations	Links	Ref.
1	Lasseter R.H. (2002)	1984	0	[40]
2	Caragliu A. (2011)	1489	2	[21]
3	Albino V. (2015)	1217	3	[60]
4	Batty M. (2012)	961	1	[61]
5	Mancarella P. (2014)	750	1	[50]
6	Walling R.A. (2008)	734	1	[136]
7	Silva B.N. (2018)	460	1	[65]
8	Raza M.Q. (2015)	428	0	[359]
9	Hatziargyriou n. (2013)	419	0	[7]
10	Alam m.r. (2012)	404	0	[53]
11	Zhou B. (2016)	385	0	[224]
12	Coster E.J. (2011)	308	0	[2]
13	Lazaroiu G.C. (2012)	297	2	[64]
14	Barnes M. (2007)	277	0	[429]
15	Hernandez L. (2014)	266	0	[357]
16	Khator S.K. (1997)	266	0	[24]
17	Ustun T.S. (2011)	265	0	[430]
18	Asmus P. (2010)	244	1	[12]
19	Frade I. (2011)	235	0	[333]
20	Calvillo C.F. (2016)	229	0	[365]
21	Farzin H. (2016)	224	0	[252]
22	Gamarra C. (2015)	219	0	[173]
23	Khodaei A. (2015)	215	1	[172]
24	Lotfi H. (2017)	213	1	[175]
25	Che L. (2014)	208	1	[432]
26	Ghayvat H. (2015)	198	0	[361]
27	Arefifar S.A. (2012)	190	1	[191]
28	Bui V.H. (2018)	183	0	[253]
29	Gopiya Naik S. (2013)	182	0	[110]
30	Arefifar S.A. (2013)	177	2	[192]
31	Pesaran H.A M. (2017)	172	1	[165]
32	Gangale F. (2013)	169	0	[439]
33	Che L. (2017)	156	2	[169]
34	Yin C.T. (2015)	151	1	[63]

. The bibliographic coupling of 26 sources whose work is cited in other sources is given in Figure 38. The heat map of top 176 documents that have received the most citations is shown in Figure 39 while their bibliographic coupling is shown in Figure 40. Similarly, the heat map of co-occurrence of all searched keywords and their relationships is shown in Figure 41.

17. Conclusions

In this paper, smart distribution mechanisms are discussed with respect to planning perspectives. The need for active planning for distribution mechanisms is discussed briefly. Each smart distribution mechanism is discussed individually with conceptual frameworks and working principles, applications, and motives. This paper covered literature of the last 10 years in the field of planning of smart distribution networks.

The planning of each smart distribution network is classified on the basis of general planning perspectives, generation and load forecasting, resource allocation, and topology selection. Generation and load forecasting facilitate generally medium-term and long-term planning. Long-term forecasting of generation and demand facilitates long-term contracts for energy purchasing and trading in the electricity wholesale market. Moreover, generation and load forecasting offers deployment of a wide range of demand-side management applications in real time and day-ahead scenarios. Forecasting in smart homes and smart cities refers to the prediction of climatic conditions, human behaviors, and energy consumption by appliances to facilitate reduction in overall energy consumption and maximization of user comfort.

Resource allocation refers to identifying the optimal type, size, and location for deployment of distributed generation, compensation devices, switching and other protection devices. Optimal resource allocation includes the minimization and minimization if certain objectives, such as maximizing profit mand minimizing energy cost and power losses, etc. Resource allocation in smart homes and smart cities is referred to optimal deployment of sensors, HVAC systems, and other cooling and heating sensing devices, which refer to cost and comfort optimization of the residents. Optimal topology selection of any distribution network structure refers to reconfiguration of an existing system in order to enhance its capabilities for large REG penetration adoption and train it to cope with other electrical, operational, and technical challenges in the case of any abnormal contingencies.

In the next section, different policies and on-ground projects of developed and developing nations are summarized. These demonstrate the global focus on adoption of active distribution networks and clean and sustainable energy production and consumption. In the last sections, research gaps in the existing literature, challenges, and issues in adoption of these technologies are presented. Recommendations based on these challenges and issues and faced in the literature are presented. Adoption of and research into new clean sustainable hydrogen economy-based storage infrastructures, developing new planning paradigms and frameworks, and leapfrogging techniques for developing nations are presented briefly.