Economic Optimal Coordinated Dispatch of Power for Community Users Considering Shared Energy Storage and Demand Response under Blockchain

Abstract

In recent years, user-side energy storage has begun to develop. At the same time, independent energy storage stations are gradually being commercialized. The user side puts shared energy storage under coordinated operation, which becomes a new energy utilization scheme. To solve the many challenges that arise from this scenario, this paper proposes a community power coordinated dispatching model based on blockchain technology that considers shared energy storage and demand response. First of all, this paper analyzes the operating architecture of a community coordinated dispatching system under blockchain. Combined with the electricity consumption mode of communities using a shared energy storage station service, the interactive operation mechanism and system framework of block chain for coordinated dispatching are designed. Secondly, with the goal of minimizing the total cost of coordinated operation of the community alliance, an optimal dispatching model is established according to the relevant constraints, such as the community demand response, shared energy storage system operation and so on. Thirdly, the blockchain application scheme of community coordinated dispatching is designed, including the incentive mechanism based on the improved Shapley value allocation coordination cost, and the consensus algorithm based on the change rate of users’ electricity utilization utility function. Finally, the simulation results show that the proposed community coordinated dispatching strategy in this paper can effectively reduce the economic cost, reduce the pressure on the power grid, and promote the consumption of clean energy. The combination of the designed cost allocation and other methods with blockchain technology solves the trust problem and promotes the innovation of the power dispatching mode. This study can provide some references for the application of blockchain technology in user-side energy storage and shared energy storage.

1. Introduction

In September 2020, China proposed to have CO₂ emissions peak by 2030 and achieve carbon neutrality by 2060 [1]. Constructing a new power system that includes photovoltaic power generation, wind power generation, and other new energy sources is an important step toward carbon peaking and carbon neutrality [2]. Meanwhile, as technological innovation and the Energy Internet have advanced, cutting-edge information technologies such as Big Data, Artificial Intelligence, Cloud Computing and Blockchain have become widely integrated into the generation, transmission, storage, conversion, and consumption of energy systems [3]. The mutual coupling of information flow and energy flow to obtain the maximization of value flow [4,5] has become the new goal of intelligent energy system dispatching. The traditional mode of social electricity utilization has been changed. Users participate in the demand response and adjust the energy utilization plan, which can reduce the electricity utilization cost while meeting the energy demand. In this process, advanced digital technology supports more complex market mechanisms and trading modes, and the user has the dual identity of power user and power regulator. Many community user groups equipped with photovoltaic and wind power generation systems have emerged due to China’s electricity market reform, new energy development and access to user-side distributed power generation devices [6]. Multiple communities can form an alliance for coordinated dispatching nearby to meet users’ electricity consumption in the region while interacting with energy and information. Communities make a demand response (DR) based on grid pricing and their circumstances to achieve renewable energy consumption and enjoy economic benefits [7]. However, this consortium will necessitate more flexibility in electric energy dispatching.

Energy storage technologies are widely used in frequency regulation, peak regulation, smoothing out fluctuations in renewable energy output, demand-side response, and improving user reliability [8]. They constitute a key technology in supporting the development of the Energy Internet and new power systems. In addition, new energy storage technologies such as new lithium-ion batteries, compressed air, hydrogen (ammonia) energy storage, and thermal (cold) energy storage are giving impetus to green development [9]. Using an energy storage system to store electric energy during low-price periods and supply electricity to users during peak-price periods can save users money while relieving grid peak regulation pressure. National and local governments have made efforts to promote the application of energy storage technology, and its development prospects are broad. A smaller energy storage capacity is frequently required by home users, small communities, and small-scale commercial users. They are more price sensitive to energy storage equipment. Finding energy storage capacity modules that precisely match their load curves on the market is also more challenging. Users’ individually configured energy storage systems have a low equipment utilization rate, resulting in idle or insufficient energy storage capacity and high operation and maintenance costs [10]. These factors severely limit the widespread adoption of user-side energy storage. The energy-sharing economy has snowballed as a result of computer technology. The natural cohabitation of energy storage technology and sharing ideas has accelerated the development of shared energy storage [11]. The concept of “shared energy storage” (SES) was first proposed in China in 2018, and refers to centralized large-scale independent energy storage stations invested in and built by third parties; it is a commercial operation mode of energy storage [12]. Since 2021, guidelines for energy storage construction have been issued in Shandong, Hunan, Zhejiang, Inner Mongolia, and other provinces to encourage investment in the construction of shared energy storage stations. On 21 March 2022, the National Development and Reform Commission (NDRC) and the National Energy Administration (NEA) released the “14th Five-Year Plan” [9] for the Development of New Energy Storage, proposing to explore and promote the shared energy storage model. According to incomplete statistics, as of April 2023, among 22.2 GW/53.8 GW·h new energy storage demonstration projects released by Chinese provinces, independent or shared energy storage projects, have reached 20.0 GW/47.4 GW·h, accounting for 92% of the power scale. On the user side, the use of shared energy storage can capitalize on the differences and complementarities of different users’ load curves, carry out demand response driven by dynamic tariff mechanisms, improve the utilization rate of energy storage equipment, promote renewable energy consumption, and reduce users’ economic costs through coordinated and optimized dispatch [13], hence realizing economic, social, and environmental value creation.

Traditional renewable energy trading is mainly managed by centralized mechanisms. Each node needs to upload a large amount of real-time information to the central node of the trading and integrate and release information in real time through the centralized structure, which leads to high operating costs, poor anti-risk ability, information opacity, and other problems. As shared energy storage and users who can meet demand response participate in the electricity market as emerging energy agents, on the one hand, distributed power and data communication will occur frequently due to a large number of agents and a more complex supply-demand relationship. On the other hand, some distributed power trading is gradually coming to show the characteristics of free access to each unit, point-to-point interaction, and decentralized coordination. Therefore, more digital and intelligent technologies are needed to support the above dispatching and trading processes. More digital and intelligent technologies can be used to support the dispatching and trading process [14]. Blockchain, as a new type of database in the context of the energy Internet, is characterized by decentralization, data traceability, and transaction transparency, which can support the transparent, efficient, and orderly management of the multi-user coordinated optimization dispatching process, and is conducive to optimizing the transaction process and solving problems such as trust between transaction subjects [15]. Blockchain incorporates technologies such as asymmetric encryption, consensus mechanisms, chained blocks, and smart contracts, which can provide a solution to achieve energy, information, and value coordination and sharing among communities [16]. Consequently, we consider shared energy storage and demand response in this paper, which is facilitated by blockchain technology, to study the optimal coordinated dispatching strategy among communities to minimize economic operating costs.

The main contributions of this paper are as follows. (1) Based on the user-side energy storage scenario, the dispatching strategy of communities is studied. Based on the service model of a new energy storage station and shared energy storage, a more economical energy use model to reduce the cost of electricity and promote the consumption of renewable energy is proposed. (2) Blockchain technology is combined with community power coordinated dispatching to fully develop its advantages such as security and scalability. The blockchain application scheme and operation process are designed in terms of smart contract, incentive mechanism, consensus algorithm, block data structure, etc. (3) An optimal dispatching model for communities considering shared energy storage and demand response is built to guide communities’ decisions, for example, pertaining to the demand response of each community, the power purchased from the grid, and the charging and discharging behavior using shared energy storage. The model is nested in a smart contract to enable the search for optimal policies. This paper uses a case study to verify the validity of the proposed method and provide a reference for the application of user-side energy storage.

The remainder of this paper is organized as follows. Section 2 reviews relevant research. Section 3 analyzes the coordinated dispatching architecture of communities based on blockchain. Section 4 establishes an optimal dispatching model. Section 5 discusses a design for the blockchain application scheme. Section 6 carries out a case study. Section 7 concludes the paper.

2. Literature Review

Power system dispatching involving large-scale forms of renewable energy, such as wind power and photovoltaic energy, has acquired new characteristics. The combination of flexible resources such as demand response, energy storage and renewable energy resources also promotes renewable energy elecrticity generation and low-carbon process [17]. The research surrounding power system dispatching operation focuses on power-side dispatching [18], distribution network dispatching [19], user-side dispatching [20,21], and micro-grid dispatching [22,23]. The optimal dispatching objectives mainly include minimal economic operating costs [18,19,21,22,23,24,25], minimum carbon emissions [17,24], maximum system profit [20], maximum power supply reliability, and so on. For instance, Ahsan et al. (2020) studied an optimization model, which is formulated as a mixed-integer linear programming problem and is solved in ILOG optimization studio with CPLEX solver. Furthermore, it solved the problem of optimized power dispatch for the solar photo-voltaic-storage system with multiple buildings in bilateral contracts [20]. The optimal dispatching results show that, combined with the time-of-use price (TOU), the installation of residential photovoltaic and energy storage can achieve higher profits. Lu et al. (2020) proposed an community energy hub optimal load dispatching model aimed at reducing the total costs of the community energy hub, including the cost of operation and CO2 emission of the system [21].

At present, the high investment cost of energy storage limits the development and application of user-side energy storage. To solve this problem, sharing economy can be introduced into user-side energy storage technology. Walker and Kwon (2021) took the use of shared energy storage and demand response in residential communities into account at the same time to studied how to assign shared energy storage to the residential consumers and how to control changing and discharging of the shared energy storage by multiple consumers [26]. Scholars have carried out relevant studies on the application scenario, pricing mechanism [27], service mode [28], capacity configuration [29], optimized operation [30], and income allocation [31,32] of user-side shared energy storage. Compared with the self-built shared energy storage system, users have better independence and flexibility when using the energy storage invested and maintained by the shared energy storage station operator [33], and they can change the electricity consumption plan and join or exit at any time [28]. At the same time, China is vigorously promoting new energy storage entities to participate in the power market as independent operators, and large-scale energy storage stations can also bear higher risks.

Although the economy and development prospects of shared energy storage have been verified, the traditional centralized energy trading mode of centralized control, dispatching and bidding is not only inefficient, but also difficult to ensure the security of trading information [34], and does not conform to the characteristics of “openness, equivalence, interconnection and sharing” of Energy Internet. The actual community electricity dispatching is more suitable for adopting the decentralized mechanism [35,36,37,38]. For example, He et al. (2021) proposed a peer-to-peer (P2P) community energy trading framework that one stakeholder is responsible for installing, connecting, managing, and maintaining the specific P2P sharing network and possesses a publicly accessible battery (available) energy storage system [39]. Blockchain is a new technology that integrates distributed data storage, point-to-point transmission, an encryption mechanism, etc. As a distributed, decentralized and information-secure network database system, blockchain has natural matching advantages to combine with distributed power trading among multi-communities with shared energy storage [40]. Blockchain technology can support frequent communication between and within multiple communities and encrypt transmissions. In addition, it has similar topological form with the interconnection system of multi-communities and shared energy storage, and has strong extensibility and compatibility, which is difficult to realize with centralized dispatching.

As a distributed and decentralized network database system, there have been some discussions on the application scenarios of blockchain in the energy field [41]. Wang et al. (2022) reviewed popular application scenarios of energy blockchain and analyzed the generic limitations of blockchain and their impacts on energy systems [42]. Gong et al. designed and analyzed various levels and related application scenarios based on energy blockchain [43]. At the same time, there is some research on using blockchain technology to solve problems such as power dispatching and trading. Based on the smart contract of blockchain, Wang et al. (2022) designed the cloud service platform for the integrated energy market and implemented decentralized intelligent dispatch [44]. Damisa and Nwulu (2022) proposed an Ethereum smart contract to provide residential storage capacities for shared facility controllers [45]. The above literature provides a reference for the application of blockchain technology in power dispatching, but it is limited to combining the advantages of blockchain technology with the traditional energy trading process, and does not consider the situation of the continuous development of distributed energy, demand response, shared energy storage and other systems in the community micro-grid, and the deep coupling of energy/information flow. Power users have been transformed into independent actors with active choice in power economic dispatching, and complex competition and coordination exist among multiple agents [46].

From the above literature summary, it is apparent that previous studies still suffer from some shortcomings, primarily including the following three points. First of all, (1) at present, studies on shared energy storage mainly focus on energy storage systems, and few studies consider coordinated dispatching behavior and economic benefits of demand response and charging and discharging by shared energy storage from the perspective of users. (2) Traditional power dispatching is mostly based on a centralized coordinated platform, but more research should be conducted on shared, integrated, and trusted decentralized dispatching modes in the future. (3) The specific applications of trust mechanisms, smart contract, consensus algorithm, incentive method, and other contents in blockchain technology in user-coordinated economic dispatching still need to be further studied.

3. Coordinated Dispatching Model of Communities

3.1. Coordinated Dispatching Architecture

Based on blockchain technology, the operational architecture of the community coordinated dispatching system studied in this paper is shown in Figure 1.

Based on the traditional model of users purchasing electricity from the grid, we propose a user electricity consumption model based on a sharing economy that uses shared energy storage station services. The community is an electricity consumer that geographically exists in the form of a community organization. This community may contain many households and small businesses that need electricity, but since they may be equipped with distributed generation units, their identity will switch to the producer sometimes. In general, however, it is still mostly a user. Multiple communities coordinate to reduce the cost of using energy storage resources and optimize economic dispatch by regulating energy use behavior to participate in interactive responses and exploiting the complementarity of each community’s electricity use behavior. Each community can be seen as a small microgrid, complete with Distributed Generation (DG) and Load Aggregator (LA). DG, as a user-side power supply, provides power to users in communities via distributed wind power, photovoltaic power, and other micro-source devices. LA includes fixed loads, reducible loads (such as air conditioning, etc.), and transferable loads (such as washing machines, dishwashers, etc.). Under the guidance of the electricity price signal, each community adjusts the demand of electricity load according to the time of use price (TOU), and reduces the electricity utilization cost by dispatching a controllable load. Each device within a community acts as the same subject for coordinated economic dispatch and safe operation externally, and aggregates the flexible resources within its own service area internally to promote a balance between energy supply and demand. Due to the fluctuation and uncontrollability of distributed power generation output, the energy storage system is also required to charge and discharge to compensate for the fluctuation of the community’s self-built power supply so as to improve the efficiency of new energy utilization, reduce the dependence on the grid and save economic dispatching costs. In the above dispatching mode, the shared energy storage station is a vital operator providing reliable charging and discharging services for the community. Communities can use the shared energy storage charging and discharging requirements without limitations on time and capacity. Communities pay the service fee for the day settlement period, and each community has meters installed in connection with the shared energy storage station online. When the community has extra electricity, the extra electricity is deposited into the storage station. When the community is short of electricity, the needed electricity is prioritized from the storage power station. The service fee payable is calculated according to the amount of electricity stored and retrieved by the storage power station in each period. Communities predict distributed power generation output and electricity load based on historical power generation and electricity consumption data, combined with demand response, and coordinate to make plans for charging and discharging using shared energy storage stations. Using energy storage devices requires minimal operating economic costs, while the communities save the investment costs required to install and maintain energy storage devices.

Shared energy storage station operator takes advantage of capital to establish large, shared energy storage stations among communities for unified operation and management, and collects service fees from communities. By taking advantage of the differences in power consumption behavior of the same communities at different times and different communities at the same time, it is possible to maximize the investment in the least energy storage to meet the communities’ energy storage usage needs. At the same time, shared energy storage station operators can make full use of the scale effect of energy storage devices. Large-scale shared energy storage stations have lower unit costs than communities investing in distributed energy storage, which can reduce the total investment costs of energy storage stations and shorten the payback period of energy storage devices.

3.2. Blockchain System Framework

The processes of community coordination and dispatching necessitate access to multiple communities’ equipment power operation information (DG, LA, etc.), and frequent distributed power and data communication will appear among communities. Blockchain data storage management technology can store community interaction data at a low cost and with high efficiency. Blockchain communication link technology’s decentralized and encrypted transmission can enable efficient, accurate, and global information interaction and sharing between individuals, as well as provide data support and decentralized management by the coordination and dispatching process and the safe operation of regional power. Blockchain and community coordinated power dispatching systems have a similar topological form, which contributes to solving the problems of poor scalability and compatibility of centralized systems. Blockchain technology uses consensus mechanism for distributed decision making, uses smart contracts safely and automatically completes the optimal dispatching operation of multiple communities; it also uses technical characteristic of incentive mechanisms to promote the active participation of each community in the system regulation response, which can realize the autonomous management of each community node in the network. From the perspective of technology integration, a blockchain interaction operation mechanism and blockchain system framework of coordinated economic power dispatching for communities is designed as shown in Figure 2.

The blockchain architecture and nodes are mapped to the operational model and functional entities of the communities coordinated dispatching, and a virtual node is set to monitor the network status of the system comprehensively. Multiple communities are used as blockchain primary nodes to connect other primary nodes with cutting-edge computing and reading capabilities, with blockchain functions such as billing and settlement, query, and verification. Primary nodes contain physical entity units such as DGs and LAs, which are dispersed in the blockchain cyberspace as sub-nodes and can process measurement of related entities, control data, and make them available to the system for sharing.

As seen in Figure 2, in the blockchain perception layer, the basic data are collected and pre-processed by arranging different types of sensors, smart meters, etc., in decentralized nodes. The data obtained from the perception layer will form a blockchain system consisting of an asset chain, commercial chain, and technology chain through blockchain technologies such as cryptography, distributed computing, consensus mechanism, smart contract, and incentive mechanism. The asset chain can realize large-scale distributed “trustless” asset management and write the information of adjustable electric energy, dispatching period, and geographical area into the blockchain after energy authentication, providing data storage and computing basis for blockchain group implementation. The commercial chain develops trading contracts. The technology chain estimates and analyzes the status of each node among the communities to realize coordinated operation, decision making, and dispatching of communities’ electricity consumption.

3.3. Coordinated Dispatching Mechanism Based on Blockchain

The launch of the shared energy storage service model is attracting small and medium-sized communities to participate in coordinated dispatching, and the trust machine of blockchain technology can empower community nodes with the right to know, the right to choose, the right to make decisions, and the right to have a voice. Firstly, communities join the blockchain as nodes through screening and verification. Secondly, a response plan is formed based on an optimized dispatching strategy pre-compiled, pre-agreed, and stored in the code of the smart contract. The specific process of optimal dispatching is as follows: (1) Data uploading stage. Each community uploads the source and load forecast results for the next 24 h, as well as information on demand responsiveness; (2) Dispatching decision stage. The optimal strategy is output according to the coordinated economic dispatch target and model; (3) Incentive regulation stage. This uses the consensus mechanism to guide incentive allocation and promote active participation of communities in coordinated dispatching. Lastly, the book-keeping node is selected according to the consensus mechanism, the required data are recorded as the content of a single block in the commercial chain, and the hash value is generated according to the rules, while the time stamp is added to record the execution time and stored in the blockchain to guide the dispatching operation of the next community’s electricity consumption behavior.

The specific coordinated optimization dispatching model and blockchain technology solution will be detailed in the following subsections.

4. Optimal Dispatching Model

Based on the above, the coordinated operation mode of communities under the block chain includes two types of subjects: communities and shared energy storage stations. As the key node considered in this paper, communities are equipped with distributed new energy power generation and flexible power load resources. The dispatching strategy determines the total economic benefits, the renewable energy consumption level, the effect on the peak-cutting and valley-filling to grid, and so on.

4.1. Coordinated Dispatching Costs of Communities

Communities access the shared energy storage station to consider the load demand response and adjust the electricity consumption plan to determine the interaction power between it and the community shared energy storage station, the power purchased from the grid, the load reduction and transfer power, the capacity of the shared energy storage system, and the charge-discharge power. Multiple communities can be regarded as a community alliance. In this paper, day-ahead economic dispatching is carried out with the optimization goal of minimizing the total daily operation cost of the community alliance. The total daily operation cost $C$ includes the service cost paid to the energy storage station $C_1$ and the cost of purchasing electricity from the grid $C_2$, which is specifically expressed as:

$$\min C=C_1+C_2$$

(1)

$$C_1={\displaystyle \sum _{i=1}^n{\displaystyle \sum _{t=1}^T\alpha (t)\times [P_{c,i}(t)+P_{d,i}}}(t)]\times \Delta t$$

(2)

$$C_2={\displaystyle \sum _{i=1}^n{\displaystyle \sum _{t=1}^T\beta (t)\times P_{b,i}(t)}}\times \Delta t$$

(3)

In the formula: $n$ is the number of community nodes; $T$ is the time period number of dispatching; $\alpha (t)$ is the charge-discharge cost coefficient paid by communities to the energy storage station; $P_{c,i}(t)$ and $P_{d,i}(t)$ are the charging and discharging power of the $i$ community using SES during $t$ period; $\beta (t)$ is the time-of-use price of power purchased from the grid; $P_{b,i}(t)$ is the power purchased from the grid for the $i$ community during $t$ period.

4.2. Operation Constraints of Communities

Ignoring transmission loss and equipment loss, the coordinated optimal dispatching of power for communities should meet the power balance constraint of community operation, load demand response constraint, charge-discharge constraint using shared energy storage station and power purchased from the grid constraint.

(1)

Electric power balance constraint

$$P_{pv,i}(t)+P_{wind,i}(t)+P_{b,i}(t)+P_{d,i}(t)-P_{c,i}(t)=P_{l,i}(t)-P_{cut,i}(t)+P_{trans,i}(t)$$

(4)

In the formula: $P_{pv,i}(t)$ and $P_{wind,i}(t)$ represent the photovoltaic and wind power generation output of the $i$ community during $t$ period respectively; $P_{l,i}(t)$ is the predicted power of electrical load before the demand response of the $i$ community during $t$ period; $P_{cut,i}(t)$ and $P_{trans,i}(t)$ are the reducible load power and the transferable load power respectively of the $i$ community during $t$ period.

(2)

Load demand response constraints

The community-side load is divided into rigid electric load and flexible electric load. A rigid electric load has poor flexibility; it can only accept power supply in a fixed period. Flexible electrical loads are more flexible and can be reduced or transferred to achieve demand response. In order to more unified describe demand response ability of communities, we assume that $k$ is the proportion of electricity that can be reduced by communities, $m$ is the proportion of load that can be transferred by communities, and $\epsilon$ is the proportion of electricity that can be adjusted by communities at each time period. The following formula (5) is the upper and lower limits constraint of the load power that can be reduced; Formula (6) is the upper and lower limits and the equal in total constraints of the transferable load power; Formula (7) is the upper and lower limit constraint of the flexible electrical load power.

$$0\le P_{cut,i}(t)\le k\times P_{l,i}(t)$$

(5)

$$\left\{\begin{array}{c}\left|P_{trans,i}(t)\right|\le m\times P_{l,i}(t)\\ {\displaystyle \sum _{t=1}^T}{P}_{trans,i}(t)=m\times {\displaystyle \sum _{t=1}^T{P}_{l,i}(t)}\end{array}\right.$$

(6)

$$\left|P_{trans,i}(t)-P_{cut,i}(t)\right|\le \epsilon \times P_{l,i}(t)$$

(7)

(3)

Communities use shared energy storage charge-discharge power constraints

Due to the introduction of a shared energy storage station, communities can store or access a certain amount of electric energy at each time period.

$$\left\{\begin{array}{c}0\le P_{c,i}(t)\le P_{\max }\times U_{c,i}(t)\\ 0\le P_{d,i}(t)\le P_{\max }\times U_{d,i}(t)\\ U_{c,i}(t)+U_{d,i}(t)\le 1\end{array}\right.$$

(8)

In the formula, $P_{\max }$ is the maximum charge-discharge power of communities using the shared energy storage station. $U_{c,i}(t)$ and $U_{d,i}(t)$ are the charging status bits and discharging status bits of shared energy storage used by the $i$ community during $t$ period. The values are 0 or 1.

(4)

The purchasing electricity from the grid non-negative constraint

$$P_{b,i}(t)\ge 0$$

(9)

4.3. The Shared Energy Storage System Operation Constraints

The shared energy storage station in the community provides communities with charge and discharge services. The shared energy storage system at run time should consider the power balance constraint of the energy storage system, the power constraints of the energy storage charge and discharge, and the constraint of the continuity of the state of charge.

(1)

The shared energy storage system electric power balance constraint

$${\displaystyle \sum _{i=1}^n\left[P_{d,i}(t)-P_{c,i}(t)\right]}=P_d^{SES}(t)-P_c^{SES}(t)$$

(10)

In the formula, $P_c^{SES}(t)$ and $P_d^{SES}(t)$ are the charging and discharging power of the shared energy storage system respectively.

(2)

The shared energy storage system charge-discharge power constraints

$$\left\{\begin{array}{c}0\le P_c^{SES}(t)\le P_{\max }^{SES}\times U_c^{SES}(t)\\ 0\le P_d^{SES}(t)\le P_{\max }^{SES}\times U_d^{SES}(t)\\ U_c^{SES}(t)+U_d^{SES}(t)\le 1\end{array}\right.$$

(11)

In the formula, $P_{\max }^{SES}$ is the maximum charge-discharge power of the shared energy storage system; $U_c^{SES}(t)$ and $U_d^{SES}(t)$ are the charging and discharging status bits of the shared energy storage system during $t$ period. The value can be 0 or 1.

(3)

The shared energy storage system SOC continuity constraints

$$\left\{\begin{array}{c}{E}_{SOC}(t)=E_{SOC}(t-1)+[{\eta }_c{P}_c^{SES}-\frac{{\eta }_d}{P_d^{SES}}]\Delta t\\ {\varphi }_{SOC}^{\min }\times E_{\max }^{SES}\le E_{SOC}(t)\le {\varphi }_{SOC}^{\max }\times E_{\max }^{SES}\end{array}\right.$$

(12)

In the formula, $E_{SOC}(t)$ is the state of charge of the shared energy storage system during $t$ period; ${\eta }_c$ and ${\eta }_d$ represent the charging and discharging efficiency of the shared energy storage system respectively. ${\varphi }_{SOC}^{\min }$ and ${\varphi }_{SOC}^{\max }$ are the upper and lower limits of SOC respectively of the shared energy storage system. $E_{\max }^{SES}$ is the maximum capacity of the shared energy storage system. After one operating cycle, the electricity of the shared energy storage system is returned to the initial state to ensure the sustainability of the next operating cycle.

5. Application Scheme Design of Blockchain Technology

The standard layers of the traditional Bitcoin blockchain are the application layer, protocol layer, network layer, transport layer, and data layer [47]. Later, with the development of Web 3.0 and Decentralized Ledger Technology (DLT), it is believed that the main layers of blockchain include the infrastructure or hardware layer, data layer, network layer, consensus layer, and application and presentation layers [48]. Among them, the consensus layer is one of the most significant layers of the blockchain. This layer is responsible for transaction authentication. It is also referred to as a consensus mechanism and maintains the blockchain’s decentralized design. The application layer, which is the uppermost layer of blockchain, includes various applications implemented based on blockchain technology, such as digital currency, smart contracts, etc. Combined with the communities’ electricity trading scenario, we divide the technical architecture of blockchain into the following six layers: data layer, network layer, consensus layer, incentive layer, contract layer, and application layer.

It can be divided into two modules: the underlying technology and the upper layer application. Among the underlying technical modules, the data layer defines the data structure. The network layer specifies the network communication protocols such as networking mechanism, data dissemination mechanism, and data verification mechanism. The consensus layer specifies the consensus mechanism to be adopted. The incentive layer specifies the issuance and allocation mechanism of cryptocurrency. This layer motivates nodes to maintain the secure operation of the blockchain system through economic balancing. In the upper application module, the smart contract layer carries smart contracts, corresponding data states and other script codes. The application layer is the application scenario corresponding to the blockchain in this paper. This means that the community alliance considers the collaborative economic dispatching scenario of demand-response and shared energy storage. The details are shown in Figure 3.

5.1. Ethereum-Based Smart Contract Development

The basic communication method of a blockchain network is the peer-to-peer network, which is different from the traditional centralized service model of server and client. The transmission of data and messages in the blockchain peer-to-peer transmission mechanism is carried out directly between nodes, and nodes can choose to join or exit the network at any moment. This paper uses the Ethereum network to build the application framework of blockchain system. Given the similarity between the infrastructure of the Ethereum network and the blockchain, the Ethereum network has features such as easy interaction of smart contracts and mature consensus protocols. It can safely and reliably automate decentralized transactions and inspire participating subjects to jointly maintain and monitor distributed data. At the same time, the Ethereum network has the advantages of being scalable and anti-attack, which can realize the security of accessing the platform and maintaining the system by a large number of participating subjects.

A smart contract in blockchain technology is an executable code deployed on the blockchain, a computer transaction protocol that automatically executes contracts. It needs to be triggered and executed in conjunction with the optimized dispatching model and incentive mechanism constructed in this paper. Considering many complex factors such as autonomous decision making, the intelligence of each community and uncertainty in renewable energy output, the smart contract in the blockchain application scenario proposed in this paper is used to optimize the day-ahead economic dispatching of communities and guide the scheduling operation strategy of each community node. It simplifies the transaction process, improves the rationality of each subject and obtains the maximum utility within the adjustable range. The trigger condition of the smart contract is that each community node which has reached consensus has broadcasted the information of respective PV, wind power output, load forecast, and demand response ability etc., and its solution process needs to satisfy the decision model of each community coalition node. The decision variables in the above model are the reducible and transferable load power of each community considering the demand response, the charge-discharge power and charge-discharge status bit of each community using shared energy storage, the electricity purchased from the grid of each community, the capacity state of the shared energy storage system, the maximum charge-discharge power and maximum capacity, and the charge-discharge power and charge-discharge status bit of the shared energy storage station. After linearizing the nonlinear constraints by Big-M method, the community optimization dispatching model is transformed into a mixed-integer linear programming problem. The above models and methods are nested in a smart contract to achieve optimal strategy searching.

5.2. Design of Incentive Mechanism Based on Improved Shapley Value Method for Sharing Coordinated Dispatching Costs

Since each community belongs to different interest subjects, their electricity generation and consumption characteristics are different. They have different degrees of demand for the use of shared energy storage and different load demand response capabilities. The contribution degree of each community to the alliance is also different. Therefore, a reasonable cost-sharing mechanism is an essential way to achieve stable operation of the alliance. In addition, the classical blockchain technologies represented by Bitcoin and Ethereum use consensus based on proof mechanisms such as PoW (Proof of Work) and PoS (Proof of Stake), and mostly use digital currency as an incentive mechanism. After each successful mining and confirmation, a new block is generated and the publicly elected book-keeping node receives the digital currency. This process relies on the nodes’ computing power and financial power. The mining process of PoW consensus consumes a lot of power resources, and PoS consensus relies on the number of coins held by nodes. The more equity in a node, the easier it is to gain bookkeeping rights, which is more likely to cause centralization problems. Therefore, the blockchain book-keeping reward designed in this paper does not use computing power and tokens; instead, it is the economic cost saved by the community coordinated power dispatch. The incentive mechanism in this paper is based on game theory to achieve optimal cost sharing among the nodes in each community alliance, and is also based on the competitive relationship and proof process in the incentive mechanism of the existing blockchain incentive layer. This process can correspond to the process of competition for computing power, number of tokens, etc. in the traditional blockchain co-governance transaction model.

The Shapley value method is the classical method of income allocation in cooperative games. This paper is based on the improved Shapley value method to share among multiple nodes the savings resulting from coordinated dispatching after considering shared energy storage and demand response. The operating cost of each community node before coordinated dispatching minus the cost savings distributed by the improved Shapley method is the final dispatching cost of that node.

There are $(\left|S\right|-1)!$ rankings for community $i$ to participate in coalition $S$, where $\left|S\right|$ is the number of communities included in coalition $S$, and there are $(n-\left|S\right|)!$ rankings for the remaining $n-\left|S\right|$ communities. The different combinations of rankings that community $i$ participates in divided by the random combination of rankings for $n$ communities is the weight of the benefit that community $i$ should share in the whole coalition, which is $\frac{(\left|S\right|-1)!(n-\left|S\right|)!}{n!}$. The marginal contribution created by community $i$’s participation in different alliances $S$ for itself and the alliances is recorded as $v(S)-v(S\backslash \left\{i\right\})$. Then, the benefit ${\phi }_i(v)$, shared by community $i$ from the cost savings of coordinated scheduling, can be calculated as shown in Formula (13).

$${\phi }_i(v)={\displaystyle \sum _{S\subset N}\frac{(\left|S\right|-1)!(n-\left|S\right|)!}{n!}}\left(v(S)-v(S\backslash \left\{i\right\})\right)$$

(13)

In the formula, $N=\left\{1,2,..,n\right\}$. $v(S)$ is the benefit generated by the entire alliance $S$ when community $i$ participates in alliance $S$. $v(S\backslash \left\{i\right\})$ is the benefit generated by alliance $S$ when community $i$ is removed.

However, the above model has shortcomings: the marginal contribution of each community is regarded as consistent. Since the characteristics of each individual are different, the Shapley value method should be improved.

In order to promote the local consumption of renewable energy, reduce energy waste and lower the pressure on the grid, communities with higher clean energy waste rates and less impact on the grid are given higher weights in their benefit distribution. This is to balance their energy storage costs incurred for the consumption of distributed electricity supplies, and accordingly encourage them to join the blockchain alliance for coordinated dispatching. The new weight of the community $i$ is calculated as $R_i$ according to the size of the consensus function. The sum of the weights of all communities is 1. The difference $\Delta R$ between the new weight $R_i$ of community $i$ and the old weight is shown in Formula (14).

$$\Delta R_i=R_i-1/n$$

(14)

The sum of the weight differences of all communities is 0. The apportionment results of the conventional Shapley value method can be fine-tuned using $\Delta R_i$. The fine-tuned amount $\Delta {\phi }_i(v)$ is given in Formula (15).

$$\Delta {\phi }_i(v)=\Delta R_i(S)$$

(15)

The new allocation result for community $i$ is shown in Formula (16).

$${\phi }_i{}^{\ast }(v)={\phi }_i(v)-\Delta {\phi }_i(v)$$

(16)

In the formula, ${\phi }_i{}^{\ast }(v)$ is the cost saving of the allocation of community $i$ after considering the improved Shapley value method.

The proposed incentive mechanism based on the improved Shapley value method to share the cost of coordinated dispatching can encourage communities to actively join the blockchain alliance. This can distribute the decision-making right and a voice to the nodes in the alliance. Moreover, any malicious behavior of any node in the alliance eventually leads to a dramatic reduction in its own benefit after multiple coordinated blockchain-based dispatching operations. To ensure that their benefits are optimal in the long term, the nodes in the alliance must comply with and maintain the alliance incentives, which in turn effectively solves the trust problem among nodes.

5.3. Consensus Algorithm Based on the Power Utility Ratio Function

The key to achieving an operating environment of blockchain network to supervise and check each other is to use consensus algorithms to ensure the viability and security of the decentralized network, so that the system can be maintained even when nodes do not enforce rules or are offline. Considering that there is no token in the incentive layer designed in this paper, the goal is to achieve the optimization of the decision-making model of each node, encourage the consumption of clean energy and reduce energy waste, reduce the pressure on the grid, and establish a multi-trust dispatching environment as the goal. Thus, a consensus algorithm based on the power utility ratio function $\Delta r(i)$ of communities is designed. The node with the largest $\Delta r(i)$ obtains the book-keeping right and records the block.

Communities hope to reduce the cost of electricity as far as possible while obtaining a reliable and high-quality electric energy supply, and obtain better service at the same time, so the concept of power utility is put forward [49]. The utility defined in this paper is the sum of communities’ satisfaction with the new energy generation consumption and the reduction in power grid purchasing after using the dispatching mode in this paper. The power utility function is described in the form of quadratic function [50], as shown in Formula (14). Power utility $r(i)$ is represented by:

$$r(i)={\displaystyle \sum _{t=1}^T\left\{v{[Z_i(t)]}^2+u{Z}_i(t)+w{[X_i(t)]}^2+z{X}_i(t)\right\}}$$

(17)

In the formula, $Z_i$ is the power of wind or light waste when community $i$ does not use shared energy storage. $X_i$ is the reduced power of electricity purchased from the grid after community $i$ considers shared energy storage and demand response. $v,u,w,z$ are the parameters of communities’ power utility function, which can reflect communities’ demand preference for energy and influence the size of demand. In order to intuitively and simply study the influence of energy waste rate and grid power purchase reduction ratio, the parameters are set as follows: $v=w=0$,$u=z=1$. At this time, the ratio of power utility $\Delta r(i)$ can be expressed as:

$$\Delta r(i)=\frac{{\displaystyle \sum _{t=1}^T\left\{Z_i(t)\right\}}}{{\displaystyle \sum _{t=1}^T\left\{P_{pv,i}(t)+P_{wind,i}(t)\right\}}}+\frac{{\displaystyle \sum _{t=1}^T\left\{X_i(t)\right\}}}{{\displaystyle \sum _{t=1}^T\left\{P_{b,i}(t)\right\}}}$$

(18)

The higher $\Delta r(i)$ is, the better it supports the coordinated dispatch mode under the blockchain, and the easier it is to obtain the blockchain bookkeeping right. In view of the defects of the current blockchain consensus algorithm, if $r(i)$ is chosen as the consensus algorithm, the network discourse power may be easily controlled by large communities. To sum up, this paper proposes using the power utility function ratio $\Delta r(i)$ as the blockchain consensus algorithm. Whether a node can obtain the bookkeeping right mainly depends on its own sustainable and healthy development.

5.4. Block Data Structure and Operation Process of Block Chain System

After the execution of a coordinated optimization dispatch, the community node with the largest value completes the bookkeeping and stamps the block time, which proves that all transactions are valid and ensures the traceability of all transactions after the event. The blockchain data structure is shown in Figure 4. The block body mainly contains the dispatching results of the current day and the electricity consumption plan of each community in the next day, including the charge-discharge power using shared energy storage, the electric power purchased from the grid, the load reduction and transfer power, the operation condition of the shared energy storage system, community operation cost and so on. At the same time, the power utility function of each community is calculated to prepare the next block consensus algorithm. All the above data are converted into binary Merkle roots by the hash algorithm and stored in the block head to ensure the privacy of the data. Based on the above blockchain node decision model, incentive mechanism, smart contract-solving algorithm and consensus algorithm, the major operation process of the blockchain node in this paper can be obtained, as shown in Figure 5.

6. Case Study

This section verifies the rationality and validity of the community transaction mode and coordinated dispatching model under the proposed blockchain through a case study. The case includes two communities with different distributed electricity supplies and one community without electricity supplies, which are represented as Community 1, Community 2, and Community 3. Set the total number of dispatching periods within one day to 24. Community 1 is equipped with photovoltaic power generation devices, Community 2 is equipped with photovoltaic and wind power devices, and Community 3 is a small, conventional load user. The typical daily load prediction of each community and the output of new energy units are shown in Figure 6. The initial SOC of the shared energy storage station is 0.3, while the maximum and minimum charge states are 0.9 and 0.1 respectively, the charging and discharging efficiency is 0.95, and the service fee is 0.33 yuan/kWh. The maximum charging and discharging power of each community using the shared energy storage station is 1000 kW. The time-of-use price of the grid is shown in

Table 1. Time of use price parameter.

	Period	Price (Yuan)
Peak	09:00–13:00, 19:00–23:00	1.12
Parity	07:00–09:00, 13:00–19:00	0.72
Valley	00:00–07:00, 23:00–24:00	0.36

. The adjustable proportion of communities’ flexible load each period is 0.2, the proportion of the reduced electricity is 0.1, and the proportion of the transferred load is 0.15. In Matlab 2016, the commercial solver Cplex 12.8 and the Yalmip toolbox were used to solve the problem.

6.1. Optimal Result

The cases in this paper are assumed to be carried out under the premise of accurate prediction and absolute rationality of all participants. Considering the coordinated dispatching mode of shared energy storage and demand response based on the above blockchain, the above scenarios are optimized. First, Community nodes 1, 2, and 3 broadcast their own distributed power output curve, load curve, demand response capability, energy loss and other information respectively. Other nodes on the energy blockchain, such as the grid and shared energy storage station, publish relevant parameters such as TOU price and shared energy storage service fee. Secondly, after all the information is included in the block, the smart contract is triggered to obtain the optimal coordinated dispatching strategies of shared energy storage station charging and discharging, load response and others of communities in the future 24 h, and the information is returned to the blockchain as the next dispatching plan.

Figure 7, Figure 8 and Figure 9 shows the optimal dispatching results of the three communities connected to the shared energy storage station. Each dispatching period is 1 h. Figure 10 shows the adjusted load curve of each communities after demand response. The interactive electricity movement between the shared energy storage station and communities in each community is shown in Figure 11. When the power value is negative, it indicates that the energy storage station is charging, and the community uses the energy storage station to store energy. If the power value is positive, it means that the shared energy storage station is discharging, providing electric energy for the communities. The configuration result of the shared energy storage station is that the capacity is 1699.35 kW·h, and the maximum charge-discharge power is 470 kW. The charge-discharge power and electricity state of the energy storage station are shown in Figure 12.

As shown in Figure 7, Community 1 can meet the load demand by purchasing electricity from external grid during 0:00–8:00, 16:00–19:00, and 23:00–24:00 because the photovoltaic power output cannot meet the load demand of communities. From 9:00 to 16:00, the photovoltaic output is greater than the load demand of the community, and the remaining electric energy inside the community is stored through the shared energy storage station to avoid electricity waste. During the period from 13:00 to 14:00, photovoltaic output reaches the maximum value of 560 kW, and Community 1 uses the shared energy storage station to charge and reach the maximum power of 428 kW. During 19:00–23:00, photovoltaic output is less than the community’s load demand. Since this period mainly belongs to the peak period of grid price, in order to minimize the operationg cost, Community 1 uses the discharge of the shared energy storage station to meet the load demand. From 21:00 to 22:00, Community 1 uses the discharge of the shared energy storage station to meet the maximum power, 112 kW.

As shown in Figure 8, the load and wind power output of Community 2 fluctuated greatly. During 0:00–7:00 and 19:00–21:00, Community 2 mainly obtains electricity from wind generator units and the grid. Community 2 mainly obtains electric energy from wind generator units, photovoltaic generator units and the grid during 7:00–8:00 and 16:00–18:00. From 8:00 to 9:00, electric energy is obtained from wind generator units, photovoltaic generator units, shared energy storage stations and the grid. During 9:00–12:00, 15:00–16:00, 18:00–19:00, electric energy will be obtained from photovoltaic generator units, wind generator units and shared energy storage stations. From 19:00 to 21:00, the electric energy is mainly obtained from wind generator units and shared energy storage stations. The maximum discharge power of the shared energy storage reaches 210 kW during 19:00–20:00. From 21:00 to 24:00, the wind generator units will produce energy. Community 2 uses the shared energy storage service to store excess wind power during the periods of 2:00–3:00, 12:00–15:00 and 21:00–24:00. The maximum charging power reaches 82 kW at 13:00–14:00.

As shown in Figure 9, since Community 3 is not equipped with distributed generator units, it needs to purchase electricity from the grid and use shared energy storage system discharges to supplement the electric energy. In the periods of 0:00–9:00, 16:00–19:00 and 23:00–24:00, communities mainly reduce energy costs through load response regulation and purchasing electricity from the grid. Most of these periods are in the valley stage and the parity stage, and the price of purchasing electricity from the grid is low. In other periods, the cost of purchasing electricity from the grid is relatively high. Community 3 purchases electricity from the shared energy storage station. During 10:00–11:00 and 15:00–16:00, Community 3 uses the shared energy storage station to reach the maximum discharge power of 60 kW.

Figure 10 shows the demand response strategies of each community. The initial load of each community electricity plan is mainly concentrated in the peak electricity price hours. Under the premise of meeting the total electricity consumption demand and the electric quantity constraint of each period, the communities can adjust the electricity demand. Communities can reduce the load from 8:00 to 24:00 or shift the load to 0:00–8:00 to implement demand response, reducing the cost of purchasing electricity during peak electricity price times.

As shown in Figure 11, Community 1 mainly uses shared energy storage stations during 9:00–16:00 to store excess photovoltaic electricity. The shared energy storage stations discharge is used to supplement the electric energy during 19:00–23:00 when the grid purchase price is high. Due to the strong volatility of wind power, Community 2 selects shared energy storage stations for charging and discharging at different periods during the day to ensure load demand. Community 3 is not equipped with distributed electricity supply, and uses energy mainly in peak periods and parity periods of grid price from 9:00 to 16:00 and 19:00 to 23:00. So, Community 3 chooses to use the shared energy storage station to purchase electricity. The interaction between each community and the shared energy storage station is consistent with the analysis results in Figure 7, Figure 8 and Figure 9.

As shown in Figure 12, the shared energy storage station is in a charging state during the period of 2:00–3:00, 10:00–15:00, and 23:00–24:00, and is in a discharging state during the period of 8:00–10:00, 15:00–16:00, and 18:00–23:00. It reaches the maximum charging power 470 kW during 13:00–14:00, and reaches the maximum discharge power 370 kW during 19:00–20:00. After discharge, the electricity of the shared energy storage station reaches the minimum value $0.1E_{\max }^{SES}$ from 9:00 to 10:00. From 14:00 to 15:00, the electricity of the shared energy storage station goes up to the maximum value $0.9E_{\max }^{SES}$. Finally, after discharge, it is restored to the initial state $0.2E_{\max }^{SES}$ at 23:00–24:00. After a cycle of operation, the shared energy storage station returns to the initial state SOC to ensure the normal operation of the next stage. In addition, it can be seen from Figure 7, Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12 that the electric load of communities in each community reaches a state of balance, and there is no phenomenon of light waste or wind waste. It is conducive to the consumption of new energy.

6.2. Economic Analysis

6.2.1. Comparison of Dispatching Strategies for Different Scenarios

The following 3 scenarios are set up to compare and analyze the economics of different dispatching strategies. The results are shown in

Table 2. Optimization scheduling results of Scenario 1.

Community	Operating Cost (Yuan)	Electricity Purchased from the Grid (kW·h)	Wasteful Electricity (kW·h)
Community 1	1224.40	1840	1390
Community 2	2089.80	2710	230
Community 3	1114	1515	0
Total	4428.20	6065	1620

,

Table 3. Optimization scheduling results of Scenario 2.

Community	Operating Cost (Yuan)	Electricity Purchased from the Grid (kW·h)	Shared Energy Storage Capacity (kW·h)/Power (kW)	Energy Storage Cost (Yuan)
Community 1	1319.70	1380	-	-
Community 2	1704.69	2178.82	-	-
Community 3	726.90	1025	-	-
Total	3751.29	4583.82	1348.68/480	1296.51

and

Table 4. Optimization scheduling results of Scenario 3.

Community	Operating Cost (Yuan)	Electricity Purchased from the Grid (kW·h)	Shared Energy Storage Capacity (kW·h)/Power (kW)	Energy Storage Cost (Yuan)
Community 1	-	1271.50	-	-
Community 2	-	1147.51	-	-
Community 3	-	795.50	-	-
Total	2876.56	3214.51	1699.35/470	1297.23

.

Scenario 1: Communities do not use energy storage systems, and they don’t make demand response.

Scenario 2: Communities are only connected to energy storage systems without making demand response.

Scenario 3: Communities consider coordinated dispatching of shared energy storage and demand response. That is, the alliance dispatching strategy with blockchain.

It can be seen that communities can effectively reduce their dispatch costs by choosing to use the charging and discharging services of shared energy storage stations. When each community does not use the energy storage system to consume distributed generation, its total dispatching operation cost is 4428.20 CNY. When communities connect to the shared energy storage station to consume wind power and PV, the total community operating cost is 3751.29 CNY, with a 15.29% reduction.

In Scenario 1, the community purchases all of the shortfall from the grid when the renewable energy generation power (wind and PV) is less than the electric load demand. When the renewable energy generation power is greater than the electric load demand, electricity is wasted, resulting in waste of renewable energy. The community operating cost of Scenario 1 is the cost of purchased electricity from the grid. From Table 2, we can see that the waste phenomenon is more serious in Community 1. The total electricity waste in Community 1 and Community 2 reaches 1620 kW·h.

In Scenario 2, each community is connected to a shared energy storage station to consume wind power and PV. The communities’ operation strategy is uniformly deployed by the energy storage station, and the optimization goal is to minimize the total daily operation cost of the communities; the dispatching results are shown in Table 3. Comparing with Table 2, it can be seen that when communities connect to the shared energy storage station, the total amount of electricity purchased from the grid is reduced by 24.42%, which reduces the pressure on the grid and promotes the consumption of renewable energy. The configuration capacity of the shared energy storage station in Scenario 2 is 1348.69 kW·h, the maximum charge/discharge power is 480 kW, and the revenue of the energy storage station is 1296.51 CNY.

Scenario 3 is a scenario showcasing the coordinated optimization dispatching strategy designed in this paper. Based on the blockchain transaction model, the community alliance considers shared energy storage and demand response to coordinated dispatching, and the optimization goal is to minimize the cost of coordinated daily operation of the community alliance, and the dispatching results are shown in Table 4. Comparing with Table 2 and Table 3, it can be seen that the coordinated dispatching strategy designed in this paper can save more costs. Compared with Scenario 1, the total amount of electricity purchased from the grid is 47% lower at this time. The revenue of the shared energy storage station is 1297.23 CNY.

6.2.2. Costing for Each Community under the Blockchain

The optimization objective of Scenario 3 in Section 6.2.1 is to minimize the total cost of coordinated operation in the three communities. However, in the blockchain-based dispatching process, there is also billing and settlement of the dispatching cost for each community. Based on the incentive mechanism of the blockchain mentioned above, the daily running cost of each community is calculated by the improved Shapley value method. This process needs to take into account the running costs of various other forms of alliances first. Set up Scenarios 4–9 as follows.

Scenario 4: Community 1 and Community 2 choose blockchain-based coordinated dispatching, and Community 3 only accesses shared energy storage.

Scenario 5: Community 1 and Community 3 choose blockchain-based coordinated dispatching, and Community 2 only accesses shared energy storage.

Scenario 6: Community 2 and Community 3 choose blockchain-based coordinated dispatching, and Community 1 only accesses shared energy storage.

Scenario 7: Community 1 chooses blockchain-based coordinated dispatching, and Community 2 and Community 3 only access shared energy storage.

Scenario 8: Community 2 chooses blockchain-based coordinated dispatching, and Community 1 and Community 3 only access shared energy storage.

Scenario 9: Community 3 chooses blockchain-based coordinated dispatching, and Community 1 and Community 2 only access shared energy storage.

Table 5. Optimization scheduling results of Scenario 3–9.

Scenario	Alliance Form	Total Operating Cost (Yuan)	Cost Saving (Yuan)
3	{Community 1, Community 2, Community 3}	2876.56	874.73
4	{Community 1, Community 2}	3005.11	746.18
5	{Community 1, Community 3}	3323.02	428.27
6	{Community 2, Community 3}	3034.01	717.28
7	{Community 1}	3484.88	266.40
8	{Community 2}	3162.56	588.73
9	{Community 3}	3585.50	165.80

shows the total operating costs and cost savings for coordinated dispatching of community alliances based on the Shapley value method for various rankings.

Table 6. Allocation results of cost based on conventional Shapley value methods.

Community	Operating Cost of Scenario 2 (Yuan)	Cost Saving Allocation (Yuan)	Operating Cost of Scenario 3 (Yuan)
Community 1	1319.70	211.27	1108.43
Community 2	1704.69	516.94	1187.75
Community 3	726.90	146.52	580.38
Total	3751.29	874.73	2876.56

and

Table 7. Allocation results of cost based on improved Shapley value methods.

Community	Operating Cost of Scenario 2 (Yuan)	Improvement Consideration		The Improved Shapley Value Method
		Wasteful Power Rate	Decline Rate of Electricity Purchased from the Grid	Cost Saving Allocation (Yuan)	Operating Cost of Scenario 3 (Yuan)
Community 1	1319.70	52.45%	30.90%	296.82	1022.88
Community 2	1704.69	6.15%	57.66%	510.64	1194.05
Community 3	726.90	0	47.49%	67.28	659.63

show the allocation results of the coordinated dispatching costs for each community based on the conventional and improved Shapley value method, respectively.

As shown in Table 7, the daily operating costs of Community 1, Community 2, and Community 3 are all significantly lower than those in Scenario 1 and Scenario 2. Compared to Scenario 1, the daily operating costs of the three community settlements decreased by 16.46%, 42.86%, and 40.79% respectively. Compared with scenario 2, they decreased by 22.49%, 29.95%, and 9.25% respectively. It can be seen that the optimization strategy of forming a coordinated dispatching alliance among communities is highly economical. Combined with the results in 6.2.1, it can be demonstrated that the method takes into account the overall and individual benefits and achieves the desired goal.

6.3. Blockchain Operation Analysis

The feasibility and economy of the proposed optimal dispatching and cost sharing methods are verified in the previous subsection. In the following, the above algorithms are combined with the blockchain operation process to analyze their rationality.

In the case of this paper, 3 communities are used as blockchain nodes to analyze the blockchain operation process for a typical day. After triggering the incentive mechanism and smart contract, the dispatching strategy and operation cost are obtained, and the power utility ratio function $\Delta r(i)$ is obtained for each community. All three community nodes have the right to compete for bookkeeping. When the book-keeping rights are determined based on the power utility $r(i)$, Community 1 and Community 2 have a relatively higher probability of being awarded bookkeeping due to their larger power and load sizes and higher demand response capabilities. In this paper, we choose to use $\Delta r(i)$ as the blockchain consensus algorithm, and the consensus function values and their weights are shown in

Table 8. Consensus function values and their weights.

Community	Power Utility Ratio Rate		Power Utility
	$\mathbf{\Delta }\mathit{r}(\mathit{i})$	Weight	$\mathit{r}(\mathit{i})$	Weight
Community 1	0.843	0.431	1958.500	0.374
Community 2	0.638	0.326	1792.486	0.343
Community 3	0.475	0.243	1481.180	0.283

below. From this, we can determine that this block is recorded by Community 1 node, and a blockchain-based community power coordinated dispatching run is completed. In comparison with using $r(i)$ as the consensus algorithm, the bookkeeping right of Community 1 is significantly increased. Combined with the analysis in 6.2, the cost when Community 1 is connected to a shared storage station in Scenario 2 is higher than the operating cost in Scenario 1. This is because the cost of using shared energy storage to consume PV is higher than the cost with light waste. In Scenario 3, after Community 1 considers the blockchain-based coordinated dispatching operation of this paper, Community 1 is incentivized to join the blockchain by cost redistribution, and the cost of Community 1 is significantly reduced. It also indicates that the incentive mechanism based on cost sharing considered in this paper is meaningful.

It can be seen that blockchain technology can be effectively integrated with the optimization method and cost sharing method of coordinated dispatching of communities. Blockchain technology has contributed to the achievement of the above functions and created a more credible, fairer, and time-sensitive environment for communities to trade electricity.

7. Conclusions

The shared energy storage station is a new method of energy storage operation that emerged with the concept of the sharing economy. This paper focuses on the coordinated dispatching operation mode of communities based on shared energy storage stations. In order to solve trust issues, such as information asymmetry and third-party dominance in the coordinated operation, blockchain technology is introduced. Moreover, the blockchain topology and multi-user shared energy storage interconnection system are highly similar in network structure. This allows for scalability and compatibility that is difficult to achieve with centralized dispatching. Firstly, blockchain-based community coordination and optimization dispatching architecture is constructed in this paper, and the interactive operation mechanism of community blockchain alliance is designed. Secondly, a community coordination optimization dispatching decision model is constructed with the optimization objective of minimizing the total cost of the coordinated alliance operation. Thirdly, the blockchain application scheme is designed in terms of smart contract, incentive mechanism, consensus algorithm, and block structure. Finally, a typical community node is selected to simulate and analyze the above blockchain-based coordinated dispatching model. The main conclusions are presented as follows:

(1)

The coordinated dispatching model for communities constructed in this paper fully considers the shared energy storage system and its own demand response. It can realize the electricity interaction with itself and with other communities, which optimizes the communities’ electricity purchase behavior and consumption mode and significantly reduces the operation cost. In the case of this paper, the total cost of optimized dispatching is reduced by 35.04% compared to the community not using shared energy storage. The optimized total cost is reduced by 23.32% compared to the community’s unified dispatch through the energy storage station only.

(2)

The incentive mechanism of blockchain combined with the idea of game theory and the use of improved Shapley value method to share the cost of coordinated dispatching can encourage the community nodes to join the blockchain. Each community’s operating costs are significantly reduced after optimized dispatching. The coordinated dispatching under blockchain balances individual benefits with overall benefits. It contributes to effectively lowering the electricity cost for communities, promoting clean energy consumption, and reducing the pressure on the power grid. It also promotes the development of user-side energy storage, new independent energy storage and shared energy storage and improves economic efficiency.

(3)

The blockchain application schemes such as incentive mechanism, consensus algorithm, and operation process proposed for the dispatching scenario presented in the paper can be integrated with user electricity transactions. Thus, the validity of the proposed method in the paper is demonstrated.

Subsequently, we will continue to optimize the underlying blockchain technology and consider diversifying the types of blockchain nodes to make it more suitable for the actual situation of a distributed power-trading mechanism.