Capacity Assessment and Analysis of Vertiports Based on Simulation

Abstract

City air traffic as a new transportation mode has gradually attracted attention in recent years which will bring endless vitality to future urban development. An objective and accurate assessment of the vertiport capacity for UAVs (Unmanned Aerial Vehicles) is the basis for implementing air traffic flow management for UAVs, which is also a prerequisite for improving the efficiency of urban airspace resources used. Firstly, new topology designs are proposed and named as connected and compact topology designs based on the existing central airport topology design. Subsequently, three modes of operation are summarized for vertiports with multiple TLOF pads: independent operation, dependent operation, and segregated operation. In the next place, the overall traffic flow of the vertiport model is established based on AnyLogic while analyzing the logic of UAV operation in three modes as mentioned above. Eventually, according to the simulation results, the vertiport operation capacity, the UAVs delay, and surface area utilization under different operation modes and topology designs are analyzed. The simulation result shows that the overall average delay time of UAVs for independent operation mode is about 100 s less than that of segregated operation and it also shows that the utilization rate of independent operation mode under central design is as high as 54.42% while the utilization rate of TLOF pads of other design is less than 50%, and its vertiport capacity is the largest, so the independent operational mode under central configuration is the optimal combination.

1. Introduction

Urban Air Mobility (UAM), as a new type of transportation, has gradually enriched people’s way of transport and its convenience and speed give the UAVs a broad application in the field of urban distribution. UAM has existed in the form of helicopter flights for decades [1,2]. Personal Air Transportation Systems (PATS) and Small Aircraft Transportation Systems (SATS) [3] may be the main modes of UAM in the future and the control of UAVs in urban air traffic [4,5,6,7] is the basis for the safe operation.

UAM development is facing many problems, such as air traffic rules [8], noise [9], safety standards [10], and technical problems in the communication and navigation of UAVs [11]. Many of them have been addressed [12], but research on vertiport capacity and surface topology have been paid little attention. In the face of increasing UAM demand, the capacity of vertiports considering topology design and operational modes have gradually been identified as a research gap.

1.1. Vertiport Topology and Operational Modes

It is impractical to study the capacity of vertiports in isolation from their topology design and operational modes, and some research has been carried out on vertiport topology. Dimensions of Touchdown and Liftoff Area (TLOF) pads, gates and taxiways have been established by official documents originating from the FAA [13], ICAO [14], and EASA [15], but there is very little documentation on the topology of vertiports. The initial vertiport topology mainly includes single-pad, satellite, linear, and pier [16]. In order to meet the demand for large-scale UAM, the perimeter, central, and disconnected topology designed by Zelinski can support the takeoff and landing of a large number of UAVs, and the advantages and disadvantages of different topology designs are analyzed by comparing the surface area utilization and operational efficiency [17]. However, his study lacked an analysis of UAVs in terms of field and queuing delays. Not only did Vascik and Hansman [18] conduct a sensitivity analysis of different topologies by means of integer planning models which provided flexibility in assessing the capacity of vertiports in different topologies, but Ahn and Hwang [19] did it as well. However, their study lacks aircraft operational time and ground congestion analysis while the theoretical operational capacity is only considered.

It is considered that the operational modes are the key issue to the vertiports. In civil aviation, runways are categorized into independent parallel approach, dependent parallel approach, independent parallel departure, and segregated parallel operation, depending on how the runway is used for approaches and departures and whether aircraft are equipped with intervals or not. In the aspect of UAM, EASA has analyzed the operational mode of a single TLOF pad equipped with multiple flight procedures [15] but has not defined an operation mode between multiple TLOF pads. The research of Vascik and Hansman for vertiport operation mode is mainly to consider the operational relationship between pads, gates, and approach or departure fix to analyze the capacity of vertiports under different operational modes and topologies, which is mainly categorized into fully independent operations, fully dependent operations and partially dependent operations [18].

1.2. Airport Capacity Assessment Methods

Airport capacity assessment is considered to be the basis for airport construction and an important indicator for assessing the operational efficiency and service capability of airports. In civil aviation, the methods of assessing airport capacity are categorized into computer simulation models [20], assessment of controller workload [21,22], mathematical models [23], and historical data mining methods [24]. The airport capacity assessment method based on simulation mainly relies on the computer to simulate the actual operation of airports and the software that is currently being used for airspace capacity assessment includes SIMMOD, TAAM, RAMS, and AIRTOP [20]. The airport capacity assessment methods based on controller workload are mainly based on human factors [21,22]. The historical data mining methods require cleaning and mining of historical airport data, which has high-quality requirements for the data [25]. The mathematical modeling method is based on modeling the main runway, taxiway, and aircraft of the airport to obtain capacity. The most classic of this type of this method is proposed by Blumstein in 1959, which assessed the single runway airport capacity based on factors such as spacing requirements, aircraft speeds and glide slope [26].

For the capacity assessment of vertiports, some research has been carried out in recent years and the capacity assessment method of vertiports based on mathematical models is the most maturely developed. Vascik and Hansman constructed a model to assess the capacity of vertiports by considering pads, gates, and staging area constraints; they also conducted a sensitivity analysis of the ground topology of vertiports [18]. Ahn and Hwang assessed the capacity of vertiports in Gimpo by constructing an integer planning model and the airport capacity envelope is plotted [19]. However, the vertiport capacity assessment model developed by Vascik and Ahn et al. does not take into account delays in airport operations. In the process of takeoff and landing, the role of a pad is equivalent to that of a runway in civil aviation airports, so there are also studies on the modeling of TLOF pads [27], which do not take into account the utilization of surface area and operational modes. There are certainly many studies that consider the capacity of vertiports in terms of algorithms for approach scheduling. Kleinbekman et al. modeled the terminal area airspace to assess airport capacity and delays within vertiports using a rolling scheduling strategy [28]. Guerreiro et al. evaluated and analyzed the vertiport capacity and throughput based on a first-come-first-served scheduling algorithm and considered a sensitivity analysis of the main airport elements such as pads [29]. However, the assessment model developed by Kleinbekman and Guerreiro et al. does not take into account critical factors such as the operational modes and the utilization of surface area. Rimjha and Trani [30] analyzed not only the capacity and delay status of vertiports but also the utilization rate of pads by constructing an airport operation logic, however, their assessment of the capacity of vertiports still lacks the consideration of the operational modes.

1.3. Research Review Summary and Our Contributions

Although vertiport topology design, operational modes, and capacity assessment have reached some achievements respectively, the vertiport capacity assessment based on the integration of topology design and operational modes has gradually been identified as a research gap. Although some research on the capacity of vertiports has been carried out, their research only considers the number of operations without topology design, operational modes, delays, the utilization of the surface area, and taxiway congestion is relatively incomplete. While some research has considered the sensitivities of topology design, the topology is reduced to the number of ground elements such as gates and pads, and does not consider the ground layout and its interrelationships. In addition, there is still a lack of simulation platforms for vertiports. Before the construction of large-scale vertiports, simulation experiments on vertiports can reasonably grasp the service capacity of vertiports as well as the sustainable development of resources, which is considered to be the basis of airport construction.

In order to address the above problems, a complete vertiport capacity assessment method based on simulation is presented. The main contributions of this paper are as follows:

(1)

Based on the existing central airport topology design, new topology designs are proposed and named connected and compact topology designs.

(2)

Based on the multi-runway operation modes of civil aviation, the operation modes of vertiports with multiple TLOF pads are summarized and classified as independent, dependent, and segregated operations.

(3)

The operation procedures of the vertiport are constructed and the vertiport is simulated based on AnyLogic according to the topology designs and operational modes. Based on the simulation model, vertiport capacity, aircraft delays, and surface area utilization are analyzed and used to evaluate the topology and operational modes of vertiports to obtain the optimal combination of topology design and operational modes.

2. Research Approach

Figure 1 displays the approach used in this paper to present the vertiport elements and improve and design the ground topology of a vertiport, and the operational modes of a vertiport are also summarized. Based on the operation procedures of vertiports, a simulation model of vertiports has been established. The vertiport capacity is obtained based on the simulation results, the aircraft delay situation, and the vertiport main element utilization and field congestion are analyzed.

Firstly, the existing construction regulations for the main elements of vertiports and historic research on the topology of vertiports were reviewed. Aircraft dimensions were also established to determine the physical dimensions of the pads, gates, and taxiways to establish a concept for the operation of vertiports.

The second step of this study was to design new vertiport topologies based on the vertiport operation procedure. The different vertiport operation modes were summarized based on existing research and traditional civil aviation’s classification of airport operation modes.

The third step of this study was to establish a simulation model of a vertiport. According to the topologies of vertiports, the main elements of vertiports, such as gates and taxiways, were placed, and the aircraft operation logic was designed according to the operation procedures and operation modes of vertiports. Additionally, fixed parameters such as aircraft operational speed, turnaround time, approach, and departure time were set based on existing studies [18].

The last step of this study was to analyze the operation of the vertiport based on the simulation model in the third step. The Poisson distribution [23] is utilized to analyze the number of aircraft that can be fully serviced by the airport per unit of time under different topology designs and operational modes. The operation trends are observed by continuously increasing the approaching aircraft and the maximum value is used to represent the airport capacity.

3. Vertiport Elements

The infrastructure of a vertiport should include pads, aprons, and taxiways. The size of the elements within a limited surface area determines the amount of this element and has a direct and significant impact on the operation of the vertiport. The capacity of vertiports is largely related to the number of gates and TLOF pads [18]. However, different vertiport topologies need to be considered because topologies may have an impact on aircraft taxi paths and operational procedures. In this paper, the topologies are first improved based on the existing central topology [17], which are categorized into central, connected, and compact topologies. In addition, the operation mode is also a fundamental element for vertiports with multiple TLOF pads, which determines the use of the pads, the operation procedures, and whether or not safety intervals need to be equipped. This means that the mode of operation significantly affects the operational capacity of vertiports.

3.1. Main Elements of Vertiports

The main elements of vertiports include TLOF pads, gates, and taxiways, as shown in Figure 2. In some research [18], staging stands are used for aircraft maintenance and recharging in designated areas and do not provide turnaround services. TLOF pads are areas designated for touchdown and liftoff that are used by aircraft to perform approach and departure procedures. Gates are designated areas for aircraft to complete turnaround processes such as unloading cargo, passenger boarding, and other related operations. A taxiway is a delineated area where aircraft taxi on the ground or fly close to the ground.

The FAA [13], ICAO [14], and EASA [15] have established the basic elements (pads, gates, and taxiways) dimensions of vertiports and this paper refers to EASA [15] for UAV dimensions, as shown in Figure 3. The diameter dimension D of an aircraft is expressed as the diameter of the smallest circle enclosing the projection of the aircraft in the horizontal plane and the diameter dimension W of the aircraft is expressed as the total width of the aircraft projection in the horizontal plane.

Based on the existing regulations [13,14,15] and the summary of the main elements of vertiports by Preis [16], the main elements of vertiports and their physical dimensions are shown in Figure 3. The dimensions of the main elements of vertiports depend on the size of the aircraft which requires a certain safety area around it and is usually designed by selecting the maximum size of the aircraft that can be landed on the vertiport.

3.2. Topology Design

The topology of a vertiport refers to the way the pads, gates, and taxiways are connected and laid out, which directly affects the operational status and serviceability of vertiports. In this paper, connected and compact structures are proposed on top of the existing central topologies [17]. In Figure 4, the green squares indicate TLOF pads, the blue circles indicate different gates and the yellow areas indicate the taxiways. The vertiport designed in this paper requires a minimum width of 22 m, with the gate radius set at 3 m, the width of the TLOF pad set at 4 m, and the width of the taxiways set at 3.5 m. The number of TLOF pads affects the capacity of the vertiport, and changing the number of pads will directly change the vertiport capacity. Four pads are chosen for this paper, it is because the four corners of the square are symmetrical and the choice of four pads is convenient for pairing different operational modes. However, the topology design in this paper is based on the central type topology designed by Zelinski [17] with four pads, and its ground area utilization rate is also more reasonable. Meanwhile, in the research of Vascik and Hansman [18], the vertiport capacity of one TLOF pad with eight gates reaches the maximum value, so it is reasonable to choose four TLOF pads in this paper.

In the central type topology design, the distance between two neighboring TLOF pads is 18 m and each TLOF pad is equipped with six gates. For the selected TLOF pad, there is only one taxi path available for aircraft and the taxi paths of different aircraft may conflict, causing congestion on the field. Therefore, under such a topology, taxiways may become a major factor limiting the capacity of vertiports.

In the connected topology design, each TLOF pad is equipped with seven pads and the distance between two neighboring TLOF pads is 16 m, which is shorter compared to the central topology. The connected topology is designed with four main taxiways connecting the TLOF pads in sequence and the gates are evenly distributed on both sides of the taxiways so that there are multiple paths for aircraft to reach the designated TLOF pads, which is not easy to cause congestion on the field. However, compared to the central type, the longer aircraft taxi path may cause some degree of delay.

The spacing between adjacent pads is further reduced to 10 m and each pad is equipped with eight gates for the compact topology design. The effect of the reduced spacing of the TLOF pads is that aircraft on adjacent pads may interact with each other and aircraft on adjacent pads may be required to be equipped with safety intervals for safety purposes, which may reduce the operational efficiency of vertiports. The taxiway is located on the outer side of the TLOF pads making it a circular taxiway with good connectivity for compact topology. However, approaching and departing aircraft may be lined up near the pads, which can lead to congestion on the taxiways.

3.3. Operational Modes

For vertiports with multiple pads, the neighboring pads cannot meet a certain distance requirement due to the limitation of terrain or surface area, so there may be the influence of the downwash airflow between aircraft on the neighboring pads. Therefore, it is necessary to set up the corresponding operation mode for the vertiports with multiple pads.

The categorization of operational modes proposed by Vascik and Hansman is based on the relationship between the gates, pads, and approach or departure fix, which are classified as fully independent operations, fully dependent operations, and partially dependent operations [18]. However, in the topology mentioned in this paper, the access between the gates and TLOF pads is bi-directional, so only the relationship between the TLOF pads and the approach or departure fix needs to be considered. Based on the use of TLOF pads and whether aircraft on adjacent pads are required to be equipped with safety intervals, the operational modes of vertiports are categorized into independent, dependent, and segregated operations, as shown in Figure 5.

Where $d$ represents the center distance between two TLOF pads, $S_0$ represents the minimum separation that does not affect the aerodynamics and related interference of all aircraft, $\Delta s$ represents the distance safety interval, and $\Delta t$ represents the time interval.

The independent operational mode mentioned in this article is like the independent parallel approach mode in civil aviation, aircrafts on adjacent pads can perform takeoffs and landings at the same time and do not need to be equipped with safety intervals as the spacing of TLOF pads is sufficient. However, if the distance between neighboring pads is close due to the constraints of terrain and surface area, it is necessary to equip a safety spacing between aircraft on neighboring pads, thus the dependent operation is proposed. For some vertiports, different pads are used only for approaching or departing in order to make the overall direction of aircraft flow easy to manage and this mode of operation is defined as segregated operation.

4. Simulation of Vertiports Based on AnyLogic

4.1. UAVs Simulation Avoidance Principle

The simulation of the daily operation of UAVs is carried out based on the pedestrian library model of AnyLogic. The avoidance principle and path selection of AnyLogic’s built-in pedestrian library model is based on the vector sum of the target gravitational force of the UAVs, obstacle drag, and other aircraft repulsive forces, which we call the social force model, as shown in Figure 6. The social force model is based on the comprehensive judgment of the target and obstacles, which is in line with the operation conditions of UAVs in actual operation.

4.2. Vertiport Simulation Topology Settings

4.2.1. Vertiport Ground Topology Settings

In Section 3.2, this paper summarizes the vertiport ground topology classified as central, connected, and compact topology designs and the dimensions of the main elements are established (

Table 1. Basic parameters of the vertiport.

Settings	Dimensions
TLOF pad	4 × 4 m
Gate	R = 3 m
Taxiway	W = 3.5 m
Vertiport	22 × 22 m
Spacing of TLOF pads (central structure)	18 m
Spacing of TLOF pads (connected structure)	16 m
Spacing of TLOF pads (compact structure)	10 m

), which are plotted in AnyLogic at a scale of 20 pixels/meter, as shown in Figure 7.

4.2.2. Vertiport Aerial Topology Settings

In this paper, the final approach procedure and the initial departure procedure of the UAVs are simplified as a vertical descent from a vertical altitude over the landing pad or a vertical ascent from the ground to a vertical altitude over the takeoff pad. In Figure 8, during the approach process, the UAVs arrive near the edge of the holding pattern airspace (blue straight line) and directly pass through the holding pattern airspace to reach the final approach fix to complete the final process if there is no waiting queue.

Since the related concept of holding pattern airspace has been presented in this section, the following modeling assumptions about holding pattern airspace and other related assumptions are made:

(1)

The UAV queue is infinite long;

(2)

The air operation and ground gliding of UAVs are uniform motion;

(3)

The UAV will not change the TLOF pad after determining the TLOF pads;

(4)

All UAVs will request to take off after ground turnover;

(5)

The endurance of UAV is unlimited.

4.3. Simulation Operation Procedures

The main elements of finite capacity need to be constrained during simulation to ensure safe aircraft operation. The occupied status of gates, taxiways, and TLOF pads is a major determinant of UAVs path selection and the status of these major elements is updated when they are occupied. For the simulation of different operational modes and ground topologies, the operation process is more or less the same, with the difference existing in the optional pads for approach and departure.

4.3.1. Major Elements State Parameter Settings

During the modeling process, the grids are meshed for pads, gates, and taxiways. Once an aircraft exists in this grid, this grid is assigned a value of 0. The assignment of the grid determines whether or not the aircraft is released for the previous operational session and the decision to release implies that there are remaining positions.

For TLOF pads, there may be different states of availability under different modes of operation, and the state of each pad is constantly changing over time. Two binary codes are used in the simulation model to indicate the availability status of the pads, with the first bit indicating the availability of landing services and the second bit indicating the availability of takeoff services, as shown in

Table 2. TLOF pads status parameters.

Service Status	Landing Service	Takeoff Service
Occupied	0	0
Idle (Takeoff only)	0	1
Idle (Landing only)	1	0
Idle (Takeoff and landing)	1	1

.

For gates, it is only necessary to distinguish whether they are occupied or not. However, considering the long turnaround time of the UAV at the gates, the UAV should choose to taxi to the entrance of the gates that are released first and wait for it when all the gates are occupied, so the waiting time for release is added as another state quantity of the gates. Therefore, two binary codes are also used in the simulation model to indicate the availability status of the gates, as shown in

Table 3. Gates status parameters.

Quantity of State	Meaning
Available state 0	Occupied
Available state 1	Idle
Waiting time for release	Duration when the available status becomes 1

.

For taxiways, the UAV paths are gridded and the idle grids are assigned a value of 1, and the grids with existing UAVs gliding are assigned a value of 0. The real-time state of the taxiway can thus be digitally described, and each UAV searches the taxiway path while taxiing on the ground.

4.3.2. Operational Procedures

It can be seen from Table 1 that the spacing of the pads for each topology design is different. In this paper, the criterion for distinguishing different operational modes is based on the maximum airborne safety spacing of UAVs proposed by Zou [10] et al. The central and connected topologies are designed to meet the requirements $d\ge S_0$ of UAVs, and the pads are independent of each other, so there is no need for dependent operational modes between the UAVs. It is due to the spacing of the pads $d<S_0$ in the compact configuration that the pads interact with each other, and therefore the UAVs must operate in a dependent mode.

Regardless of the modes of operation, the general logic for UAVs remains the same. The traffic follows a Poisson distribution, emerges from the final approach fix corresponding to each pad in the terminal airspace, descends in altitude, and arrives at the landing pad. Subsequently, UAVs select the nearest free gate for the turnaround according to the path-finding algorithm. After a certain period of time, UAVs select the appropriate pad to perform the takeoff task and climb altitude to the initial departure fix according to the operation rules (Figure 9). In Figure 9, +1 means an increase of 1 in the number of aircraft in a queue and −1 means a decrease of 1 in the number of aircraft in a queue.

For UAVs, there are a variety of situations that may be encountered while taxiing on the ground. In the independent mode of operation, the pad may be used for both take-offs and landings and the departure aircraft waits in a waiting area around the pad until the approaching aircraft is released from the pad. Similarly, if the departure aircraft enters the pad, the approaching aircraft waits in the holding pattern airspace. For aircraft taxiing on the ground, if there is a plurality of aircraft ready to complete the departure procedure, the first aircraft to complete the turnaround selects the take-off pad and is prioritized to arrange the taxi path, and other aircraft ready for departure waits until the first aircraft passes through its apron before taxiing out in turn. The departure aircraft arrives at the area around the take-off pad for determination, and if the pad is not occupied, the departure procedure will be completed directly. If the approaching aircraft does not find any free gate when selecting, it will taxi to the waiting area around the first aircraft to complete a turnaround and waits for the aircraft to taxi out. If the taxiway of the approach and departure aircraft conflict, the departure aircraft waits at the gate until the departure aircraft passes. These aircraft taxi path conflicts can cause some delays and are a critical factor in reducing TLOF pad throughput.

4.4. Simulation Parameter Settings

Referring to the existing research [18], the fixed operational parameters of vertiports are set for the design characteristics of vertiports and operational modes. The average UAV ground taxi speed $\overline{v}$ is the speed of the aircraft taxiing on the ground taxiway and the speed of the aircraft satisfies a normal distribution with a mean value of 1 m/s. The length of the final leg descent and the length of the beginning leg ascent $H$ are consistent as mentioned in Section 4.2.1. The UAVs ground taxi safety interval $\overline{d}$ is the interval that should be maintained between aircraft in the direction of aircraft operation. The average occupancy time of landing pads $\overline{t_p^a}$ and the average occupancy time of takeoff pads $\overline{t_p^d}$ are set to ensure that aircraft have sufficient safety intervals during approach and departure. Unit time is the number of completed operations and aircraft delays at vertiports that are counted for a certain period of time during the simulation and analysis process. Turnaround time refers to the time it takes for an aircraft to complete turnaround actions such as unloading or charging on the gates, and in some studies [18] new areas such as staging stands are defined for charging and other operations. The specific values of the above relevant parameters are shown in

Table 4. Operation parameters of vertiport.

Parameters	Numerical Value
Length of last descent section $H$	50 m
Average ground taxiing speed $\overline{v}$	1 m/s
Ground taxiing safety interval $\overline{d}$	3 m
Average occupation time of landing pad $\overline{t_p^a}$	45 s
Average occupation time of takeoff pad $\overline{t_p^d}$	30 s
Unit time $T_{unit}$	1 h
Turnaround time $T_{turnaround}$	300 s

.

5. Analysis of Simulation Results

For vertiports with different topologies and operational modes, this paper evaluates their operations per unit of time. By continuously increasing the number of approaching UAVs, operations per unit of time are observed in the vertiports, and vertiport capacity is assessed by finding the maximum number of operations. Vertiport capacity can be considered as the throughput per unit of time of a vertiport, and for the definition of throughput it can be considered as the sum of the number of approach and departure flights of a vertiport [18]. However, some research considers the number of aircraft that complete the entire process of approaching, taxiing, turnaround, and performing a departure as one throughput or operation [16]. In this paper, the number of aircraft completing a complete process will be used as the throughput and its maximum value will be the vertiport capacity. In addition to this, delays for UAVs under different operational modes, the number of approaching UAVs, and topology designs are considered. Finally, the utilization and efficiency of the main elements on the ground are considered to lay the foundation for a rational and scientific selection of the topology and mode of operation of the vertiports.

5.1. Vertiport Capacity

In order to determine the capacity of vertiports under different operational modes and topologies, it is necessary to observe the operations that can be carried out by vertiports per unit of time. The Poisson distribution will be used in this paper to model the distribution pattern of approaching aircraft at vertiports. When the number of approaching aircraft is at a low level, the number of operations at vertiports may be small.

Vertiport capacity is determined by finding the maximum number of vertiport operations. In this paper, we set the parameters $\lambda$ for independent operational mode and segregated operational mode as 60, 120, 180, 240, and 300, where $\lambda =60$ means the average of the number of approaching aircraft at vertiports in one hour is 60. Each combination is simulated three times and the average value is taken. The results of the simulation are shown in

Table 5. The number of vertiport operations with different elements and the number of approaches.

	$\lambda$	60	120	180	240	300
Elements
Central-Independent		46	73	74	82	81
Central-Segregated		38	68	69	70	69
Connected-Independent		34	57	54	53	57
Connected-Segregated		41	53	53	55	53
Compact-Independent		38	42	47	46	47
Compact-Segregated		39	48	48	48	46

.

The data in Table 5 represents only the number of vertiport operations under different topology designs and modes of operation, not the capacity of the vertiports. Vertiport capacity is related to the number of gates, pads, taxiways, topology designs, and mode of operation of the vertiport. The maximum value of the number of operations under the corresponding topology designs and operational modes in Table 5 reflects the maximum service capacity of the vertiports under this combination. To observe the trend, a three-dimensional histogram is plotted in Figure 10.

In Figure 10, when the central-type topology is used in an independent mode of operation, its vertiports have the maximum capacity. Compact topology designs do not differ significantly in their vertiport service capacity between independent and segregated modes of operation, and compact topologies are weaker than central and connected topology designs in terms of service capacity. While the operations of vertiports fluctuates with the number of approaching aircraft, it remains constant when increased to a certain value, which corresponds to the vertiport capacity. However, it is found that as the number of approaching aircraft increases, the number of operations decreases under certain circumstances. The reason for this phenomenon may be that the ground taxiing speed of the aircraft is normally distributed with randomness and the selection of gate location by the aircraft during the approach and departure process is also random. It is also possible that the increase in the number of approaching aircraft leads to the occurrence of congestion on the field, which leads to a decrease in the number of operations. In general, vertiport operations increase with the number of approaching aircraft and finally stabilize.

5.2. Delays of UAVs

The delay time of UAVs determines the satisfaction of customer service and can reflect the stability and service capability of an airport service from the side, so the consideration of the delay of UAVs in different situations can provide an important reference for the construction of vertiports. In this paper, we will consider delays at vertiports in terms of the mode of operation, number of approaching aircraft, and topology designs.

5.2.1. Delays in Different Modes of Operation

The mode of operation determines the use of the different pads. In independent operation, the four pads can be used for both takeoff and landing at the same time, while in segregated operation, two of the four pads are used for takeoff and two are used for landing. In order to simulate the airport aircraft delays under different operational modes, Poisson distribution is still used to simulate the distribution of approaching aircraft, and the number of approaching aircraft in independent and segregated operation modes is set at 60 aircraft per hour, and select 100 UAVs to observe the delays, as shown in Figure 11.

Figure 11 reveals that the same type of delays in the same mode are essentially the same for each pad, with larger differences between different UAVs. For the independent mode of operation, most of the UAVs have landing delays of less than 100 s, and a small number of UAVs with delays of 200 s or more are also observed. Compared to the independent mode, the landing delay in the segregated mode is significantly higher because the independent mode has four takeoff and landing pads to perform the landing task while the segregated mode has only two pads to perform the landing task. To better compare the distribution of delays, the mean value of UAVs delay time on each pad is calculated as shown in Figure 12 and Figure 13, where the capital letters L and T denote the landing or takeoff pads, respectively, and the number after the letter denotes the pad serial number. For example, L3 indicates the third landing pad.

By comparing Figure 12 and Figure 13, it is found that the mean and standard deviation of the delay on each pad in the independent operation mode are close to each other, whereas the delay length in the segregated operation shows an obvious imbalance and the landing delay is much larger than the takeoff delay in the segregated operation. Meanwhile, since the average takeoff time of the UAV is shorter compared to the average landing time, the UAV can update the status of the pads more quickly after arriving at the pad, which makes the average takeoff delay time shorter than the landing time.

5.2.2. Delays for Different Approach Traffic Flows

In the operation of vertiports, quantities of UAVs approach and departure scenarios are often encountered and the ability of vertiports to tolerate these UAVs to complete the operation of the vertiports demonstrates the stability of the vertiport services, so the study of the delay of vertiports by the large traffic flow is of great significance. In order to analyze the effect of larger traffic flows on airport delays, the number of approaching aircraft at each vertiport is expanded to 120 per hour. The delays of vertiports facing larger traffic flows are investigated for independent and segregated operational modes under the central configuration, as shown in Figure 14 and Figure 15.

From Figure 14 and Figure 15, it can be seen that with the increase in the number of approaching aircraft, the takeoff and landing delays of UAVs are significantly increased, regardless of whether the independent or segregated operation mode is used. Compared with the takeoff delay, the growth of the landing delay is significantly larger than the takeoff delay and the landing may become a key node limiting the service capacity of the airport and specific growth rate is shown in

Table 6. Delay growth rate of each landing TLOF pad.

Delayed Growth Rate	TLOF Pad ①	TLOF Pad ②	TLOF Pad ③	TLOF Pad ④	Average Value
Independent landing	2863.2%	326.8%	656.3%	472.9%	1079.80%
Independent takeoff	610.2%	255.6%	66.7%	437.3%	342.45%
Segregated landing	754.3%	2503.2%	-	-	1628.75%
Segregated takeoff	-	-	66.7%	66.7%	66.70%

.

From Table 6, it can be seen that when the number of approaching aircraft increases from 60 to 120 per hour, the delays in the independent and segregated operation modes show different characteristics. For both operation modes, landing delays increase sharply, which will lead to a decrease in vertiport service competence in the subsequent time period. Although the growth rate of takeoff delays for the segregated mode of operation is not large, its growth rate of landing delays is too large, which may hinder the normal operation of vertiports. However, the independent mode of operation is more balanced.

Comparing Figure 11 and Figure 16, it is found that the delays of the takeoff and landing pads under the independent operational mode with a small number of approaching aircraft are relatively small and stable. However, as the number of aircraft increases, some pads under the independent operational mode start to accumulate delays such as TLOF pad 1. Despite the cumulative delays, the delays of pads remain around a certain level and do not increase explosively. For the segregated mode of operation which is faced with a small number of approaching aircraft, it already generates cumulative delays at vertiports, and the delays at its landing pads such as TLOF pad 1 and TLOF pad 2 are accumulated and continue to grow when the number of approaching aircraft increases. In the above analysis, the independent operational mode has delays in the case of a larger number of approaching aircraft, but its delay time is kept at a certain level and controllable, while the delay time of the segregated operational mode is growing and not improved. Therefore, the independent mode of operation has better stability than the segregated mode of operation.

According to Figure 11 and Figure 16, it is found that the delays are the most serious for TLOF pad 1 regardless of whether it is an independent or segregated mode of operation. This is because when an aircraft is within the airspace of the terminal area, the occupancy status of all the pads is determined first, and if they are all in the occupancy status, then it will go to the first one by default, which results in the number of approaching aircraft for the TLOF pad 1 to be higher.

5.2.3. Delays for Different Surface Topology Designs

For different vertiport topology designs, the pads and the gates are connected in different ways, so the takeoff and landing delays and other indicators in the vertiports may be different. In the simulation process, the number of approaching aircrafts of the three topology designs is controlled at 60 per hour and the simulation of the three topology designs is carried out by using the independent operational mode and the segregated operational mode, as shown in Figure 17.

From Figure 17, it is found that the connected and compact topologies show the same delay characteristics, with higher instability and fluctuation in delay duration, while the overall delay in independent operation is smaller and the approach and departure delays in this operational mode are more balanced and stable. Additionally, approach delays in the segregated operational mode are higher while departure delays are at a lower level, showing unstable and fluctuating characteristics. However, the compact topology design adopts a lower level of safety in the independent and segregated operational modes, so the dependent operational modes are considered separately for it, as shown in Figure 18.

When vertiports adopt a compact topology and the dependent operational modes, either landing or takeoff delays are much greater than the other two topology designs. Its previous delays continue to stack up leading to continued lengthening of subsequent UAV queues, and the delays make it difficult to form a stable continuous flow and congestion is difficult to be relieved.

5.3. Surface Area Utilization

In the construction of vertiports, it is necessary to consider the utilization of resources in order to build sustainable vertiports. The rational and scientific utilization of the surface area within the limited urban airspace is the basis for the construction of vertiports. Therefore, the utilization of pads, gates, and taxiways will be analyzed in this paper.

5.3.1. TLOF Pads Utilization

In order to analyze the operation of vertiports under normal conditions, a Poisson distribution with a mean value of 60 per hour for the number of approaching aircraft is selected for simulation. The utilization rate of pads is calculated as the ratio of occupied time to total available time. The occupied time refers to the physical occupied time, which depends on the limitations of the simulation software. We would enjoy the opportunity to follow up with a study to calculate the occupied time that cannot be used for other operations. In both independent and dependent operational modes, the pads can be used for takeoff and landing. In segregated operation, 1 and 2 pads are responsible for landing and 3 and 4 are responsible for takeoff, and the results of the utilization rate are shown in

Table 7. Utilization of TLOF pads in each combination.

Facility Number	TLOF Pad 1&2			TLOF Pad 3&4			Average Value
Occupied State	Landing	Takeoff	Idle	Landing	Takeoff	Idle	Utilization Rate
Central Independent	26.51%	31.79%	41.70%	33.27%	17.27%	49.46%	54.42%
Central Segregated	62.64%	-	37.36%	-	33.24%	66.76%	47.94%
Connected Independent	18.86%	30.42%	50.71%	28.09%	8.37%	63.54%	42.88%
Connected Segregated	50.68%	-	49.32%	-	29.25%	70.75%	39.97%
Compact Independent	17.54%	24.47%	57.99%	28.48%	7.34%	64.18%	38.92%
Compact Segregated	49.13%	-	50.87%	-	27.44%	72.56%	38.29%
Compact Dependent	15.93%	18.40%	65.67%	20.12%	16.41%	63.47%	35.43%

.

In Table 7, the statistics of the operations related to the compact topology design are added for a more comprehensive analysis. As can be seen from Table 7, the utilization of both landing and takeoff pads in the independent operational mode is lower than that in the segregated operational mode, which may be caused by the fact that there are four pads sharing the landing task in the independent operational mode while there are only two pads in the segregated operational mode. Overall, the utilization of the takeoff and landing pads in the segregated operation of the UAV is 6.48% less than that in the independent operation and the independent operational mode has a higher utilization.

Comparing the utilization of pads corresponding to different topology designs in the same operational mode, it can be seen that the pad utilization of the topology design of the central type is significantly higher than that of the connected type and the independent type. Since the compact type adopts the dependent operational mode, the UAVs affect each other in the adjacent vertical takeoff and landing channels, which leads to the situation that the pads are idle but unavailable and a large amount of wastage of the surface area resources. Even when the compact topology design is combined with the independent operational mode or the segregated operational mode, its average pad utilization is still less than 40%, so the takeoff and landing pads of the compact type topology design have a low utilization rate.

5.3.2. Gates Utilization

The gates are the main area to ensure the normal turnover of UAVs and are a key element in the operation of vertiports. According to the simulation experiments in Section 5.3.1, the statistics of the gate utilization under each topology design and operational mode are shown in Figure 19.

In Figure 19, for all topology designs and operational modes, the gate utilization is below 40%, which indicates that the number of gates is sufficient and is not a limiting factor for the capacity of vertiports. Comparing the different operational modes, the gates utilization rate is lower in the segregated operational mode than in the independent operational mode. In the same operational mode, the central type has the largest gate utilization and the compact type has the smallest. Therefore, the independent mode of operation has a higher resource utilization compared to the segregated mode of operation.

Although the number of gates is different in different topology designs, the utilization rate decreases as the number of gates increases, indicating that a topology design with more gates is a waste of resources. Under the condition of limited vertiport area, too many gates are not fully utilized such as compact topology design, which indirectly reflects the low utilization of surface area under this topology.

5.3.3. Taxiway Congestion Situation

The taxiway is an important ground element connecting the gates and pads and the smooth operation of UAVs on the taxiway is the basis for maintaining the efficient operation of vertiports. For the three ground topology designs, the number of approaching aircraft is uniformly selected as 60 per hour. The simulation is carried out in independent operational mode and the UAV ground operation is recorded in the form of a heat map, as shown in Figure 20. In Figure 20, darker colors indicate a more crowded area.

Comparing the heat maps of the taxiways of the three topologies, it can be found that the central topology with the pads located in the four corners of the vertiport and fewer taxiways has higher congestion and higher utilization. Meanwhile, the heat map shows that there are still some unused gates in the connected and compact types, which is a waste of resources, so the central topology is the optimal choice in terms of resource utilization.

6. Conclusions

In this paper, based on the existing vertiport topology designs, two new vertiport topology designs are proposed. Additionally, three vertiport operational modes are summarized and the vertiport capacity, aircraft delay, and surface area utilization of vertiports with different topology designs are investigated based on AnyLogic in combination with the vertiport operational modes. The main conclusions are as follows:

Firstly, the operations of vertiports are counted by continuously increasing the number of approaching UAVs per hour and the maximum value is selected as the capacity of vertiports. The results show that the central topology has the maximum capacity in the independent operational mode.

Secondly, the delay of aircrafts under different operational modes, topology designs, and the number of approaching aircrafts is observed. The results show that the segregated operation has a shorter delay time for departure, but its approach delay time is longer and unstable, while the independent operational mode is more stable.

Finally, this paper considers the utilization of the main elements of vertiports. For the different modes and topology designs of pads, despite the higher utilization of departure pads in the segregated operational mode, the overall utilization of pads in the independent operational mode is higher. Meanwhile, the gate utilization is the highest in the independent operational mode, but its taxiways are more congested due to its only two main taxiways, while it is acceptable.

Although we have considered the capacity of vertiports, delays, ground elements utilization, and other indicators, we hope that future research can consider the impact of complete flight procedures on the vertiport capacity and complete the sensitivity analysis of turnaround time and taxiing time for simulation. Meanwhile, how to match the vertiport capacity with the traffic flow and how to define the vertiport capacity by using the basic traffic flow parameters is also a worthwhile research issue.