Site selection and sizing method and system for optical storage and charging micro-grid accessed to power distribution network containing flexible interconnection branches

Abstract

The invention discloses a method and a system for locating and sizing an optical storage and charging micro-grid accessed to a power distribution network containing flexible interconnection branches. On the basis of obtaining original data, determining the position of a charging station in the power distribution network by using a charging station site selection method based on electric vehicle charging demand prediction, taking the position as an alternative node for accessing the optical storage and charging micro-grid to the power distribution network, and determining photovoltaic installable capacity on the alternative node; establishing a multi-objective optimization model for site selection and energy storage constant volume of the optical storage and charging micro-grid, and taking a feasible solution of the model as an alternative scheme; and screening out an optimal scheme from the alternative schemes by using an improved approximate ideal solution sorting method, wherein the obtained optimal scheme comprises decision information such as the access position of the optical storage and charging micro-grid, the energy storage capacity and the energy storage charging and discharging power. According to the invention, the optical storage and charging micro-grid is used as a whole for site selection and sizing, so that the problems of inconsistent site selection of a photovoltaic power station and a charging station, redundant energy storage capacity configuration and the like caused by site selection according to single equipment are avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of microgrid siting and sizing for accessing distribution networks, and particularly relates to a method and system for siting and sizing a photovoltaic-storage-charging microgrid accessing a distribution network with flexible interconnection branches. Background Art

[0002] The photovoltaic-storage-charging microgrid organically integrates photovoltaic power generation, electric vehicle charging load, and energy storage system, which is an important technical route for realizing large-scale consumption of renewable energy and efficiently supplying green electric energy for electric vehicles. The flexible interconnection branch technology that can achieve load balancing, dynamic capacity increase, and fast fault transfer is increasingly applied to the transformation of traditional distribution networks, making the distribution network structure more complex and refined energy management more difficult.

[0003] The access of the photovoltaic-storage-charging microgrid to the distribution network can improve the utilization rate of photovoltaic power generation, suppress the fluctuations of photovoltaic power output and distribution network voltage. However, the access location and capacity of the photovoltaic-storage-charging microgrid have an important impact on its effectiveness. Based on the current technical conditions, when the photovoltaic-storage-charging microgrid accesses a distribution network with flexible interconnection branches, it faces technical problems of optimal siting and sizing. The existing optimal siting and sizing technologies first consider improving the power supply quality and continuity of the distribution network, and then consider improving the utilization rate of renewable energy and the economy of the microgrid. The existing technologies mostly involve aspects such as independent energy storage siting, energy storage sizing, and independent charging station siting.

[0004] (1) Siting technology of independent energy storage

[0005] The independent energy storage siting technology takes the independently configured energy storage in the distribution network as the main body, and based on technical and economic indicators such as energy storage investment cost, node voltage fluctuation of the distribution network, and photovoltaic and wind power consumption ratio, finds the best access location of the independent energy storage system in the distribution network. At present, as a siting technology for a single device, the independent energy storage siting technology lacks the analysis of the influence of photovoltaic power generation and charging load on the energy storage location. At the same time, as a siting technology for traditional distribution networks, its influence on the flexible interconnection branches in the distribution network is not fully considered.

[0006] (2) Energy storage sizing technology of microgrid

[0007] The purpose of the energy storage sizing technology in the microgrid is to solve the contradiction between the intermittency and randomness of photovoltaic power generation and the continuity of load power supply at the lowest cost. The energy storage sizing technology in the photovoltaic-storage-charging microgrid usually first generates a typical scenario set of photovoltaic power generation, and then, with the goal of minimizing the overall cost of the photovoltaic-storage-charging microgrid and meeting the basic operation constraint conditions of the microgrid and the distribution network, determines the optimal energy storage capacity. The existing energy storage sizing technologies do not consider the influence of photovoltaic power generation and charging load on the energy storage capacity, do not consider the influence of the energy storage access location on the energy storage capacity, and do not consider the influence of the power transmission loss caused by the flexible interconnection branches on determining the energy storage capacity.

[0008] (3) Independent charging station location selection technology

[0009] The existing technology can better complete the task of locating independent charging stations in the distribution network. Based on comprehensive consideration of factors such as the travel data of electric vehicles in the area, the power supply capacity of the distribution network, the available land area, and costs, the optimal location for building a charging station is determined through establishing a mathematical model and data analysis. The independent charging station location selection technology includes two aspects: charging demand simulation and optimization modeling. Among them, charging demand simulation usually estimates the charging demand according to the collected user travel data through the Monte Carlo simulation method, which can better retain the randomness of the charging load of electric vehicles. However, when locating an independent charging station, the decision variable is only the location of the charging station. When applying this technology to the location selection and capacity determination of a photovoltaic-storage-charging microgrid, there is a situation where the optimal location of the charging station is inconsistent with the determined access locations of photovoltaic and energy storage.

[0010] To sum up, the existing single-device location selection and capacity determination technology has not considered the needs of integrated location selection and capacity determination of a photovoltaic-storage-charging microgrid, insufficiently considered the characteristics of geographical concentration and mutual influence of photovoltaic, energy storage, and charging stations, insufficiently considered the influence of flexible interconnection branches in the distribution network, and lacks a method to realize the synchronous location selection and capacity determination of the three of photovoltaic, energy storage, and charging stations. Therefore, it is urgent to develop a technology for locating and sizing a photovoltaic-storage-charging microgrid connected to a distribution network with flexible interconnection branches as a whole. Summary of the Invention

[0011] To overcome the above problems existing in the prior art, the present invention discloses a method and system for locating and sizing a photovoltaic-storage-charging microgrid connected to a distribution network with flexible interconnection branches.

[0012] The specific technical solution adopted by the present invention is as follows:

[0013] A method for locating and sizing a photovoltaic-storage-charging microgrid connected to a distribution network with flexible interconnection branches includes the following steps:

[0014] 1. Obtain the original data for locating and sizing a photovoltaic-storage-charging microgrid

[0015] The original data of electric vehicle EV includes the number of EVs N EV , the battery capacity of EV, and the power consumption per unit mileage. Among them, the battery capacity of the rth EV is C r , and the power consumption per unit mileage is e r .

[0016] The original data of the distribution network includes the typical daily output of the original distributed power sources in the distribution network, the active and reactive loads of each node on a typical day. The acquisition method is as follows:

[0017] (1) Obtain the typical daily output of the original distributed power sources in the distribution network

[0018] Cluster the daily output of the original distributed power sources in the distribution network within one year into G typical scenarios, obtain the clustering center corresponding to each typical scenario and the number of days each typical scenario appears, and use the ratio of the number of days a typical scenario appears to the total number of days in a year to represent the probability of the typical scenario occurring; then, multiply the probability of each typical scenario occurring by the output data of the t-th period of the clustering center of the scenario, and then accumulate according to the number G of typical scenarios to obtain the output data of the t-th period of the typical day; finally, calculate the output data of a total of T periods within the typical day according to the above method.

[0019] (2) Obtain the active and reactive loads of the typical day of the distribution network

[0020] For the t-th period of the i-th node in the distribution network, take the average value of the active and reactive loads of this period of this node in 365 days of a year as the active and reactive loads of this node in this period of the typical day; calculate the active and reactive loads of this node in T periods within the typical day according to the above method.

[0021] Repeat the above steps N bus times to calculate the active and reactive loads of T periods of the typical day of N bus nodes in the distribution network.

[0022] 2. Determine the location of the charging station and use it as an alternative node for the optical storage charging microgrid to access the distribution network

[0023] Select N bus nodes from the total N ch nodes in the distribution network for preferential layout of charging stations, and use these N ch nodes as alternative access nodes for the optical storage charging microgrid, including the following steps.

[0024] 2.1 Establish a probability distribution model

[0025] Establish a rectangular coordinate system on the power supply area of the distribution network. Define the straight-line distance between the starting point and the ending point of the EV as the spatial distance d r between the starting point and the ending point, and the angle between the line connecting the starting point and the ending point of the EV and the horizontal axis as the driving direction θ r . The initial state of charge S r of the r-th EV follows a normal distribution, d r follows a lognormal distribution, the driving starting point O r (x0, y0) follows a uniform distribution within the power supply area of the distribution network, and θ r follows a uniform distribution.

[0026] Use the Monte Carlo method to randomly generate the starting position O r (x0, y0) of the r-th EV and the driving direction θ rand the spatial distance d between the starting and ending points r If so, the r-th EV starts from the starting point O r (x0, y0), and moves along the θ r direction, passing through the spatial distance d between the starting and ending points r to obtain the arrival position, i.e., the end point coordinates D r (x, y).

[0027] 2.2 Calculate the remaining state of charge S of the EV when it travels to point v on the path v.r

[0028] Using the Floyd algorithm for finding the shortest path, based on O r (x0, y0) and D r (x, y), obtain the driving path of the EV. The actual driving mileage corresponding to point v on its driving path is d ov.r . According to the initial state of charge S r and the power consumption per unit mileage e r , use Equation (1) to obtain the remaining state of charge S of the EV when it travels to point v v.r .

[0029]

[0030] In the formula, η is the energy equivalent coefficient, reflecting the energy loss during the start-stop process of the EV, and η = 0.95 is taken.

[0031] 2.3 Determine the EV charging demand points

[0032] When S v.r ≤ 0.3, a charging demand will be generated. Therefore, set the charging threshold as S v.r = 0.3, and set S v.r = 0.2 as the lower limit of the state of charge for the EV battery discharge.

[0033] From the initial state of charge S r , use Equation (2) to respectively obtain the actual driving mileages corresponding to S v.r = 0.3 and S v.r = 0.2:

[0034]

[0035] According to the starting position O r (x0, y0) and the actual driving mileages corresponding to S v.r = 0.3, S v.r = 0.2, determine the EV at S v.r = 0.3 and S v.r=0.2, the corresponding geographical locations are recorded as point A and point B, and the actual destination of the EV is recorded as point C. When point C is closer to the starting point than point A, the EV has no charging demand; when point C is between point A and point B, point C is determined as the charging demand point; when point C is farther from the starting point than point B, the midpoint of the line connecting points A and B is recorded as the charging demand point.

[0036] The above is the step to determine the charging demand point of the rth EV. Repeat the above steps to finally get N EV N generated by EV d Charging demand point.

[0037] 2.4 Determine the charging demand mark point

[0038] N d Clustering is performed on N EV charging demand points to obtain N ch Clusters are formed, and the centers of each cluster are used as charging demand markers. The clustering results are represented by circles in the load indicator diagram, and the centers of the circles represent charging demand markers. The diameter d of the hth circle is h Calculated by formula (3).

[0039]

[0040] Where P h is the accumulated active power of the charging load in the hth cluster circle; n is the scale.

[0041] P h As the charging load corresponding to the hth charging demand mark point.

[0042] 2.5 Determine the alternative access nodes for the photovoltaic storage and charging microgrid

[0043] The position of the hth charging demand mark point is recorded as (x h ,y h ), respectively calculate (x h ,y h ) to all N bus The distance between the hth charging demand mark point and the distribution network node closest to the hth charging demand mark point is the candidate node for the charging station corresponding to the mark point to access the distribution network. Repeat the above steps N ch times, and finally from N bus Determine N ch The candidate access nodes of charging stations are obtained, and the candidate access node set Ω of charging stations is obtained. S ,Ω S ={n1,n2,...,n Nch},Ω S Each element in Ω corresponds to the node number of the distribution network. S It can be used as a collection of alternative access nodes for the photovoltaic storage and charging microgrid.

[0044] 3. Determine the installable PV capacity of the alternative access nodes of the PV-storage-charging microgrid

[0045] Estimate Ω S The total area A of the charging pile shed, open space and public building roof at the alternative access node numbered q in Ω q.pv , and use A q.pv Multiply by the installable PV area utilization rate α q , and calculate the installable PV area A′ q.pv , then multiply the rated power generation per unit area of the PV module β by A′ q.pv , and calculate the installable PV capacity E of the q-th alternative access node q.pv . Repeat the above steps to calculate all N S Installable PV capacities of the alternative nodes in Ω ch .

[0046] 4. Determine the access location and energy storage capacity of the PV-storage-charging microgrid

[0047] 4.1 Establish an optimization model for the site selection of the PV-storage-charging microgrid and energy storage capacity determination

[0048] Taking the access location of the PV-storage-charging microgrid, the energy storage capacity and the energy storage charge-discharge power as decision variables, and aiming at the minimum voltage offset, the minimum distribution network loss and the lowest energy storage investment cost, establish a multi-objective optimization model for the site selection of the PV-storage-charging microgrid and energy storage capacity determination

[0049] Set the following constraint conditions in the optimization model

[0050] Energy storage energy balance constraint, indicating that the remaining energy storage power at the start and end of the optimization period is the same

[0051] Energy storage charge-discharge power limit, indicating that the charge and discharge power of the energy storage in each period is between the upper and lower limits of the power allowed by the energy storage

[0052] State of charge limit, indicating that the state of charge of the energy storage in each period is between the maximum and minimum state of charge allowed by the energy storage

[0053] Energy storage capacity limit, indicating that the configured energy storage capacity is between the maximum and minimum capacities allowed to access the energy storage

[0054] Node voltage limit, indicating that the voltage of each node and each period in the distribution network is between the upper and lower limits of the node voltage

[0055] Node power real-time balance, indicating that the inflow and outflow power of each node in the distribution network reaches real-time balance

[0056] The access location constraint of the photovoltaic-storage-charging microgrid indicates that the access node of the photovoltaic-storage-charging microgrid should be selected from the alternative access nodes of the photovoltaic-storage-charging microgrid determined in step 2.

[0057] 4.2 Solving the optimization model using the improved multi-objective particle swarm optimization algorithm

[0058] Solve the optimization model using the improved multi-objective particle swarm optimization algorithm. The non-linear weight ω is used to improve the optimization performance. As the number of iterations k itera increases, the value of ω continuously decreases, and ω is calculated by Equation (4).

[0059]

[0060] In the formula, a and b are auxiliary parameters for adjusting the numerical range of the inertia weight; K max is the maximum number of iterations; Z is the influence factor, and its calculation method is shown in Equation (5).

[0061]

[0062] In the formula, U ω is the flag bit. When U ω = 1, it means that the global best has not changed for 5 consecutive iterations. When U ω = 0, it means that the global best has changed at least once in 5 consecutive iterations of the particles.

[0063] During the process of solving the model using the improved particle swarm optimization algorithm, the number of particles is set to N p , and the individual best of N p particles is updated in each iteration. The individual best obtained in each iteration is used as a feasible solution. After K max iterations, a total of K max ×N p feasible solutions are finally generated, and all the feasible solutions form the feasible solution set Ω k .

[0064] During the process of solving the optimization model, according to the active power load, reactive power load, existing distributed power generation output, charging load, and energy storage charge and discharge power of N bus nodes in the distribution network, a commercial software package is called to complete the power flow calculation, and information such as the node voltage U t.i of the i-th node in the t-th time period is obtained.

[0065] 4.3 Determining the Pareto optimal solutions and using them as alternative solutions

[0066] Compare each feasible solution in the feasible solution set Ω k with all other feasible solutions in the feasible solution set one by one, and delete the feasible solutions with a dominance relationship. Finally, only the N ffeasible solutions, and these N f solutions are the Pareto optimal solutions. Define the Pareto optimal solutions as alternative solutions, and N f alternative solutions are obtained.

[0067] 4.4 Determine the optimal solution from the alternative solutions

[0068] Use the improved Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) method to determine the optimal solution for the siting and energy storage sizing of the photovoltaic-storage-charging microgrid from N f alternative solutions. Use the Manhattan distance to represent the distances between each alternative solution and the positive and negative ideal solutions.

[0069] The optimal solution for the siting and energy storage sizing of the photovoltaic-storage-charging microgrid includes the number of photovoltaic-storage-charging microgrids connected to the distribution network, the nodes of each photovoltaic-storage-charging microgrid connected to the distribution network, the capacity of the energy storage configured in each photovoltaic-storage-charging microgrid, and the 24-hour charge and discharge power of each energy storage.

[0070] 5. Determine the optimal solution for connecting the photovoltaic-storage-charging microgrid to the distribution network

[0071] After obtaining the optimal solution for the siting and energy storage sizing of the photovoltaic-storage-charging microgrid in step 4, according to the node numbers of the photovoltaic-storage-charging microgrids connected in the optimal solution, determine the charging station load and photovoltaic capacity corresponding to the node numbers from steps 2 and 3, and supplement them to the optimal solution obtained in step 4 to obtain the complete optimal solution for connecting the photovoltaic-storage-charging microgrid to the distribution network.

[0072] The beneficial effects of the present invention include:

[0073] (1) Regarding the photovoltaic-storage-charging microgrid as a whole, an optimization method for siting and sizing its connection to the distribution network with flexible interconnection branches is designed, avoiding the problem of inconsistent siting of energy storage and charging stations caused by siting according to a single device. Centralize the installation of photovoltaic, energy storage, and charging stations, reduce the power loss transmitted in the distribution network, and at the same time reduce the land occupation and the repeated construction cost of infrastructure.

[0074] (2) Consider the non-linear power flow constraints when connecting the photovoltaic-storage-charging microgrid to a complex looped distribution network with flexible interconnection branches. During the solution process of the optimization model, transform the power flow constraints into directly calling a commercial power flow software package to calculate the power flow, avoiding the difficulty of directly solving the optimization model with non-linear power flow constraints. Description of the Drawings

[0075] Figure 1 is the flow chart of the method for siting and sizing the photovoltaic-storage-charging microgrid;

[0076] Figure 2 is the typical daily output curve of the original distributed power source at node 9 of the distribution network;

[0077] Figure 3 It is the active and reactive load curves of the typical day of node 5 in the distribution network;

[0078] Figure 4 It is the indication diagram of the charging load demand in the power supply area of the distribution network;

[0079] Figure 5 It is the charge and discharge power and state of charge curves of the energy storage;

[0080] Figure 6 It is the schematic diagram of accessing the optical storage charging microgrid in the distribution network with flexible interconnection branches. Specific implementation manners

[0081] The following further details the specific implementation manners of the present invention in conjunction with the drawings and embodiments, but it does not limit the protection scope of the present invention. Any technical solutions obtained by means of equivalent replacement or equivalent transformation are within the protection scope of the present invention. The implementation process of the method for site selection and capacity determination of the optical storage charging microgrid in the distribution network with flexible interconnection branches is as shown in the attached Figure 1 figure.

[0082] Embodiment:

[0083] Taking the access of the optical storage charging microgrid in the IEEE 33-node distribution network with flexible interconnection branches after transformation as an example, the implementation process of the method is described as follows.

[0084] 1. Obtain the original data required for the site selection and capacity determination of the optical storage charging microgrid

[0085] The original data of the EVs includes the number N EV = 500 of the EVs charging in the power supply area of the distribution network, the battery capacity of each EV, and the power consumption per unit mileage. Among them, the battery capacity C r of the r-th EV = 40 kWh, and the power consumption per unit mileage e r = 0.3 kWh / km.

[0086] The original data of the IEEE 33-node distribution network includes the typical daily output of the original distributed power sources in the distribution network and the active and reactive loads of each node on the typical day. The acquisition methods are as follows:

[0087] (1) Obtain the typical daily output of the original distributed power sources in the distribution network

[0088] Taking node 9 as an example, the acquisition process is described. Collect the output data of the original distributed power source on node 9 for T = 24 time periods every day in one year. The output data on the z-th day is expressed as P pv.z = {p z,1 , p z,2 ,..., p z,T}. Using the K-means clustering method, the daily output data within one year is clustered into G = 4 typical scenarios, corresponding to 4 seasons. The output data corresponding to the clustering center of the g-th scenario is denoted as E pv.g ={e g,1 ,e g,2 ,...,e g,T}. At the same time, the number of days N g included in the g-th scenario can be obtained. Then the probability of the g-th scenario occurring is:

[0089]

[0090] The scenario clustering results obtained through the above steps are shown in Table 1.

[0091] Table 1 Number of days and occurrence probabilities included in different scenarios in the clustering results

[0092] Scenario Number of days Occurrence probability 1 78 21.4％ 2 96 26.3％ 3 84 23.0％ 4 107 29.3％

[0093] Thus, the output data of the t-th period of the typical day can be obtained as:

[0094]

[0095] Finally, the output data of T = 24 periods of the original distributed power source on the typical day at node 9 is obtained, as shown in the appendix Figure 2 .

[0096] (2) Obtain the active power load and reactive power load of the distribution network on the typical day

[0097] Taking node 5 as an example, the acquisition process is illustrated. The historical data of the active power load and reactive power load of node 5 for T = 24 periods per day within one year is obtained. Among them, the active power load data on the z-th day is denoted as P L.z ={p L.z，1 ,p L.z，2 ,...,p L.z,T}, and the reactive power load data is denoted as Q L.z.t ={q L.z.1 ,q L.z.2 ,...,q L.z.T}

[0098] Calculate the average active power load P avg,t and average reactive power load Q avg,t of the t-th period of the typical day of node 5:

[0099]

[0100] The active power load and reactive power load curves of T = 24 periods of the typical day of node 5 are obtained, as shown in the appendix Figure 3 .

[0101] Obtain the active and reactive power loads of all N = 33 nodes in the distribution network at T = 24 time periods of a typical day in the same way. bus

[0102] 2. Determine the location of the charging station and use it as an alternative node for the optical storage charging microgrid to access the distribution network

[0103] Select the nodes for preferentially arranging charging stations from the 33 nodes of the distribution network as the alternative access nodes for the optical storage charging microgrid. The specific steps are as follows.

[0104] 2.1 Establish a probability distribution model

[0105] The probability density function of the initial state of charge S of the r-th EV is: r

[0106]

[0107] In the formula, σ s is the standard deviation of S r ; u s is the mean of S r .

[0108] The probability density function of the spatial distance d between the starting and ending points of the r-th EV is: r

[0109]

[0110] In the formula, σ d is the standard deviation of d r ; u d is the mean of d r .

[0111] Establish a rectangular coordinate system on the power supply area H of the distribution network. The coordinates of any point (x, y) satisfy a < x < b and c < y < d, where a, b, c, and d are the boundaries of the area H, and a = c = -5 km, b = d = 5 km. The starting point O r (x0, y0) of the EV follows a uniform distribution in the area H, that is, x0 follows a uniform distribution of U(a, b) = U(-5 km, 5 km), and y0 follows a uniform distribution of U(c, d) = U(-5 km, 5 km).

[0112] The driving direction θ of the EV r follows a uniform distribution of U(0, 2π).

[0113] Use the Monte Carlo method to randomly generate the starting position of the r-th EV. The simulation result of a certain time is O r (x0, y0) = O r (-4 km, -4 km), driving direction θ r= 53° and the spatial distance d between the starting and ending points r = 10 km, and the end point coordinates D are calculated from Equation (12) r (x, y) = D r (2 km, 4 km).

[0114]

[0115] 2.2 Calculate the remaining state of charge S of the EV when it travels to point v on the path v.r

[0116] Using the Floyd algorithm for finding the shortest path, based on O r (-4 km, -4 km) and D r (2 km, 4 km), the driving path of the r-th EV is obtained, and the actual mileage of the driving path is 14 km.

[0117] The remaining state of charge S of the EV when it travels to point v on the path v.r , take S v.r = 0.3 and S v.r = 0.2.

[0118] 2.3 Determine the EV charging demand points

[0119] According to the initial state of charge S r = 0.36, C r = 40 kWh, e r = 0.3 kWh / km and the energy equivalence coefficient η = 0.95, calculate the actual driving mileage corresponding to S v.r = 0.3:

[0120]

[0121] Similarly, the actual driving mileage corresponding to S v.r = 0.2 is calculated to be 22.46 km.

[0122] Since the actual driving mileage d v.r = 0.3 corresponds to is 8.42 km, which is less than 14 km, the EV will have a charging demand. ov.r

[0123] According to the actual driving path and the actual driving mileage corresponding to S v.r = 0.3, S v.r = 0.2, determine that the geographical locations corresponding to S v.r = 0.3 and S v.r = 0.2 are A(-3.58 km, 4 km) and B(11.46 km, 4 km) respectively, and the actual driving end point D of the EV r r(2 km, 4 km) is denoted as point C. Point C is between point A and point B, and point C is determined as the charging demand point.

[0124] The above are the steps to determine the charging demand point of the r-th electric vehicle. Repeat the above steps to determine the charging demand points of all EVs, with a total of N d = 500 charging demand points.

[0125] 2.4 Determine the charging demand marking points

[0126] Using the K-means clustering algorithm, cluster the N d = 500 EV charging demand points to obtain N ch = 6 cluster centers. Take the cluster centers as the charging demand marking points to obtain the charging load demand indication diagram in the power distribution network supply area, as shown in the appendix Figure 4 shown. The hollow circles in the figure represent the charging demand points. Only 80 of the charging demand points are exemplified in the figure. The dashed circles represent the clustering results, and the centers of the circles represent the charging demand marking points.

[0127] Calculate the diameter d of the h-th circle in the charging load demand indication diagram h :

[0128]

[0129] P h is the cumulative active power of the charging load within the h-th cluster circle.

[0130] Similarly, the diameters of the N ch = 6 circles in the charging load demand indication diagram can be obtained.

[0131] Take P h as the charging load corresponding to the h-th charging demand marking point.

[0132] 2.5 Determine the alternative access nodes of the photovoltaic-storage-charging microgrid

[0133] Record the position of the h-th charging demand marking point as (x h , y h ) = (0.4 km, 0.5 km), and record the position of the i-th power distribution network node as (x i , y i ) = (0.2 km, 0.3 km).

[0134] Calculate the distance d between the charging demand point (x h , y h ) and the i-th power distribution network node: i :

[0135]

[0136] Record the distances from the h-th charging demand marker point to each distribution network node, and define the obtained N bus = 33 distance values as set M, and find the minimum value d min :

[0137] d min = min{M} = 0.28 km (16)

[0138] Then d min The corresponding node 7 is the alternative access node of the charging station corresponding to the h-th charging demand marker point.

[0139] Repeat the above steps. Finally, determine 6 alternative access nodes of the charging stations from N bus = 33 distribution network nodes to obtain the set Ω of alternative access nodes of the charging stations S , Ω S = {7, 13, 16, 19, 22, 28}, and each element in Ω S corresponds to the distribution network node number. Ω S can be used as the set of alternative access nodes of the PV-storage-charging microgrid.

[0140] Meanwhile, obtain the charging loads of the alternative access nodes of the PV-storage-charging microgrid corresponding to each charging marker point, as shown in Table 2.

[0141] Table 2 Charging loads of alternative nodes of the PV-storage-charging microgrid

[0142] Alternative node number Charging load / MW 7 2.43 13 0.72 16 0.81 19 0.81 22 0.81 28 0.45

[0143] 3. Determine the installable PV capacity of the alternative access nodes of the PV-storage-charging microgrid

[0144] Taking node 7 in Ω S as an example, illustrate the steps to determine the installable PV capacity of all alternative nodes. The specific steps are as follows.

[0145] Estimate the total area A of the charging pile shed, open space, and public building roofs at node 7 7.pv = 13700 m 2 . Calculate the area A' where PV can be installed at node 7 7.pv :

[0146] A' 7.pv = α7·A 7.pv = 40% × 13700 m 2 = 5480 m 2 (17)

[0147] In the formula, α7 is the utilization rate of the area where PV can be installed at node 7.

[0148] Installable Photovoltaic Capacity E of Computing Node 7 pv :

[0149] E pv = β·A′ 7.pv = 1.6×10 -4 MW / m 2 ×5480m 2 = 0.8768MW (18)

[0150] Wherein, β is the rated power generation per unit area of the photovoltaic module, and is taken as the technical average of existing products.

[0151] Repeat the above steps to calculate the installable photovoltaic capacity of all 6 alternative nodes in Ω S as shown in Table 3.

[0152] Table 3 Installable Photovoltaic Information of Alternative Nodes of Photovoltaic-Storage-Charging Microgrid

[0153]

[0154] 4. Determine the Access Location and Energy Storage Capacity of the Photovoltaic-Storage-Charging Microgrid

[0155] 4.1 Establish an Optimization Model for Site Selection and Energy Storage Capacity Determination of the Photovoltaic-Storage-Charging Microgrid

[0156] Establish a multi-objective optimization model for site selection and energy storage capacity determination of the photovoltaic-storage-charging microgrid:

[0157] minF = {f1, f2, f3} (19)

[0158] Wherein, f1, f2, and f3 respectively represent voltage deviation, distribution network power loss, and energy storage investment cost.

[0159] The first objective function:

[0160]

[0161] Wherein, U ref is the rated node voltage; U max and U min are the upper and lower limits of the node voltage, taken as 1.05 p.u. and 0.95 p.u. respectively; U i is the voltage of the i-th node, which is the average voltage of each time period at this node. The node numbered 1 is the balancing node, and its voltage amplitude is constant and does not participate in the calculation of voltage deviation.

[0162] The second objective function:

[0163]

[0164] Wherein, P gen.tis the active power input from the main grid to the distribution network during the t-th period, as well as the total active power of all photovoltaic power generation and energy storage discharge in the distribution network; P load.t represents the total active power of the loads at all nodes in the distribution network during the t-th period.

[0165] The third objective function:

[0166] In the formula, N pcsm is the number of integrated photovoltaic, energy storage, and charging microgrids connected to the distribution network; j represents the j-th integrated photovoltaic, energy storage, and charging microgrid connected to the distribution network; c ess.P and c ess.E are the investment costs per unit power and per unit capacity of the energy storage, respectively; P ess.j and E ess.j are the rated power and capacity of the j-th energy storage.

[0167] The following constraint conditions are set in the optimization model.

[0168] Energy balance constraint of the energy storage:

[0169]

[0170] In the formula, P ess.t is the power of the energy storage system during the t-th period. When the energy storage discharges, P ess.t > 0. When the energy storage charges, P ess.t < 0. When the energy storage neither discharges nor charges, P ess.t ≤ 0; Δt is the duration of each period in the optimization cycle.

[0171] Charge and discharge power limit of the energy storage:

[0172] P ess.min ≤ P ess.t ≤ P ess.max (24)

[0173] In the formula, P ess.min and P ess.max are the lower and upper limits of the power of the energy storage system, respectively.

[0174] State of charge limit:

[0175] S ess.min ≤ S ess.t ≤ S ess.max (25)

[0176] In the formula, S ess.t is the state of charge of the energy storage during the t-th period; S ess.min is the minimum state of charge allowed for the energy storage; S ess.max is the maximum state of charge allowed for the energy storage.

[0177] Energy storage capacity limit:

[0178] E ess.min ≤E ess.j ≤E ess.max (26)

[0179] Wherein, E ess.j is the capacity of the j-th energy storage configured; E ess.min is the minimum capacity of the energy storage; E ess.max is the maximum capacity of the energy storage.

[0180] Node voltage limit:

[0181] U min ≤U t.i ≤U max (27)

[0182] Real-time power balance constraint of the node:

[0183]

[0184] Wherein, P ess.t.i is the active power provided by the energy storage connected to the i-th node in the t-th time period of the distribution network to this node; P pv.t.i is the active power provided by the photovoltaic connected to the i-th node in the t-th time period to this node; P load.t.i is the active power required by the load of the i-th node in the t-th time period; The set B r represents the set of all branches connected to the i-th node in the distribution network, s represents the branch number in the set B r ; P t.s.i is the active power of the branch s in the t-th time period. When the branch power flows into the i-th node, P t.s.i >0, when the branch power flows out of the i-th node, P t.s.i <0.

[0185] The access location of the photovoltaic-storage-charging microgrid is selected from the alternative nodes determined in step 2:

[0186] Ω T ∈Ω S (29)

[0187] Wherein, Ω T is the set of access nodes of the photovoltaic-storage-charging microgrid.

[0188] 4.2 Solve the optimization model using the improved multi-objective particle swarm optimization algorithm

[0189] Set the number of particles to N p =50, update the individual extreme values of N p =50 particles each time during each iteration, and use the individual extreme values obtained in each iteration as feasible solutions. After K max= 50 iterations, and finally a total of K max × N p = 50 × 50 = 2500 feasible solutions are generated, and the set of all feasible solutions forms the feasible solution set Ω k .

[0190] During the process of solving the optimization model, according to the active power load, reactive power load, existing distributed power generation output, charging load, and energy storage charge and discharge power of N bus = 33 nodes in the distribution network, commercial software packages such as Matpower are called to complete the power flow calculation, and the node voltages of N bus = 33 nodes and T = 24 time periods are obtained.

[0191] 4.3 Determine the Pareto optimal solutions and use them as alternative solutions

[0192] Compare each feasible solution in the feasible solution set Ω k with all other feasible solutions in the feasible solution set in turn, and delete the feasible solutions with a dominance relationship.

[0193] Finally, only N f = 8 solutions that do not dominate each other pairwise are retained in the feasible solution set. These 8 solutions are the Pareto optimal solutions and are used as 8 alternative solutions.

[0194] 4.4 Determine the optimal solution from the alternative solutions

[0195] Use the improved Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) method to determine the optimal solution for the location selection of the photovoltaic-storage-charging microgrid and the energy storage capacity determination from 8 alternative solutions, including the following 5 steps.

[0196] (1) Nondimensionalization processing of the attribute values of the alternative solutions

[0197] Take the 3 objective functions as the 3 attributes of the alternative solutions for screening the optimal solution. The attribute values are the objective function values, and f m (x w ) is the m-th attribute value of the w-th alternative solution. Perform the maximum-minimum normalization processing on the attribute values to obtain N f = 8 nondimensionalized attribute values f m (x w ) as shown in Table 4.

[0198] Table 4 Nondimensionalized attribute values of the alternative solutions

[0199] Serial number Voltage deviation index Distribution network power loss Energy storage investment cost 1 0.3958 2.7744 0.4157 2 0.3659 3.0166 0.4508 3 0.3686 2.5896 0.4020 4 0.3663 2.7649 0.4664 5 0.3642 2.8710 0.4747 6 0.3684 2.6381 0.4444 7 0.3639 2.8091 0.4444 8 0.3635 2.8397 0.4809

[0200] (2) Set the positive ideal solution and negative ideal solution of each attribute

[0201] All three attributes are cost-type attributes, and the smaller the attribute value, the better. Therefore, the positive ideal solution and the negative ideal solution f′ of the m-th attribute m+ and f′ m- are respectively taken as the maximum and minimum values in the dimensionless attribute values of the m-th attribute:

[0202]

[0203] (3) Determine the weight of each attribute using the information entropy method

[0204] Using the information entropy method, the weight λ of each attribute is obtained according to the dimensionless attribute values m .

[0205] First, calculate the entropy value e of the m-th attribute m :

[0206]

[0207] In the formula, p w.m is the proportion of the m-th attribute value of the w-th alternative in the sum of the m-th attribute values of all alternatives, and is calculated using formula (34).

[0208]

[0209] Calculate the weight λ of the m-th attribute m :

[0210]

[0211] The weight λ m satisfies 0 < λ m < 1 and N a is the number of attributes.

[0212] The calculated weights of the three attributes are: λ1 = 0.0185, λ2 = 0.0556, λ3 = 0.9259.

[0213] (4) Calculate the relative distance of each alternative

[0214] The Manhattan distance is used to calculate the distances d + (x w ) and d - (x w ) between the alternative and the positive and negative ideal solutions:

[0215]

[0216] Calculate the relative distance d(x w ) of the w-th alternative:

[0217]

[0218] The relative distances of the 8 alternative solutions calculated are shown in Table 5.

[0219] Table 5 Relative Distances of Alternative Solutions

[0220] Serial number Distance from the positive ideal solution Distance from the negative ideal solution Relative distance 1 0.4835 0.4742 0.5048 2 0.5056 0.4963 0.5046 3 0.7287 0.7194 0.5032 4 0.5132 0.5039 0.5045 5 0.5239 0.5146 0.5044 6 0.4835 0.4742 0.5048 7 0.4862 0.4769 0.5048 8 0.4835 0.4742 0.5048

[0221] (5) Determine the optimal solution for the location of the photovoltaic-storage-charging microgrid and the energy storage capacity

[0222] Compare N f = 8. The relative distances of the alternative solutions are compared, and the 3rd alternative solution with the smallest relative distance is determined as the optimal solution for the location of the photovoltaic-storage-charging microgrid and the energy storage capacity.

[0223] The optimal solution for the location of the photovoltaic-storage-charging microgrid and the energy storage capacity includes the number of photovoltaic-storage-charging microgrids connected to the distribution network, the nodes of each photovoltaic-storage-charging microgrid connected to the distribution network, the energy storage capacity configured in each photovoltaic-storage-charging microgrid, and the 24-hour charge-discharge power of each energy storage.

[0224] 5. Determine the optimal solution for the connection of the photovoltaic-storage-charging microgrid to the distribution network

[0225] After obtaining the optimal solution for the location of the photovoltaic-storage-charging microgrid and the energy storage capacity in Step 4, according to the node numbers of the photovoltaic-storage-charging microgrid connected in the optimal solution, the charging station load and photovoltaic capacity corresponding to the node numbers are determined from Steps 2 and 3, and they are supplemented into the optimal solution obtained in Step 4 to obtain the complete optimal solution for the connection of the photovoltaic-storage-charging microgrid to the distribution network.

[0226] In the optimal solution, the number of photovoltaic-storage-charging microgrids N pcsm = 2. The 24-hour charge-discharge power and state of charge of each energy storage are as shown in the appendix Figure 5 shown, and the optimal solution for the connection of the photovoltaic-storage-charging microgrid to the distribution network is shown in Table 6. The appendix Figure 6 shows the specific locations of the photovoltaic-storage-charging microgrid connected to the distribution network.

[0227] Table 6 Optimal Solution for the Connection of the Photovoltaic-Storage-Charging Microgrid to the Distribution Network

[0228]

[0229] As shown above, only the preferred embodiments of the present invention are presented, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for site selection and capacity determination of a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches, characterized in that: The following steps are involved: Step 1: Obtain the original data of the electric vehicle EV and the original data of the distribution network; Step 2: Use the charging station site selection method based on EV charging demand prediction to determine the alternative access node of the charging station, and use it as the alternative node for the photovoltaic storage and charging microgrid to access the distribution network; Step 3: Calculate the installable photovoltaic capacity of the alternative access node of the photovoltaic storage and charging microgrid determined in step 2; Step 4: Taking the access location of the photovoltaic storage and charging microgrid, the capacity of the energy storage, and the charging and discharging power of the energy storage as decision variables, a multi-objective optimization model for the site selection and energy storage sizing of the photovoltaic storage and charging microgrid is established; the optimization model is solved using the improved multi-objective particle swarm algorithm to obtain a feasible solution set; the Pareto optimal solution is selected from the feasible solution set and used as an alternative solution; the optimal solution for the site selection and energy storage sizing of the photovoltaic storage and charging microgrid is determined from the alternative solutions using the improved approximate ideal solution sorting method; Step 5: Add the charging station load and photovoltaic capacity configuration scheme determined in steps 2 and 3 to the optimal scheme for site selection and energy storage capacity of the photovoltaic storage and charging microgrid determined in step 4 to obtain the optimal scheme for connecting the photovoltaic storage and charging microgrid to the distribution network.

2. The method for site selection and capacity determination of a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches according to claim 1, characterized in that: The original data of the distribution network in step 1 includes the typical daily output of the original distributed generation, and the calculation process is: Cluster the single-day power generation of the original distributed power generation in a year into G typical scenarios, obtain the cluster center and the number of days of occurrence of each typical scenario, and use the ratio of the number of days of occurrence of the typical scenario to the total number of days in a year to represent the probability of occurrence of the typical scenario; Multiply the probability of occurrence of each typical scenario by the output data of the tth period of the cluster center of the scenario, and then accumulate the output data of the tth period of the typical day according to the number of typical scenarios G; finally, calculate the output data of a total of T periods in the typical day according to the above method; The original data of the distribution network also includes the active load and reactive load of the nodes on a typical day. The calculation process is as follows: For the tth time period, the active load and reactive load of this time period of 365 days in a year are averaged as the active load and reactive load of this time period on a typical day; the active load and reactive load of a total of T time periods in a typical day are calculated according to the above method.

3. The method for site selection and capacity determination of a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches according to claim 1, characterized in that: The specific process of charging station site selection based on EV charging demand forecast in step 2 is as follows: Obtain the original data such as the number of EVs, the battery capacity of EVs, and the power consumption per unit mileage; establish the probability distribution model of random variables such as the initial state of charge of EVs, the starting point of travel, the travel direction, and the spatial distance between the starting and ending points, and use the Monte Carlo method to generate the values ​​of the above random variables to calculate the EV travel end point; The shortest path method is then used to determine the EV's driving path. Based on the EV's initial state of charge and power consumption per unit mileage, the mileage and location when the state of charge reaches the charging threshold are calculated. This location is the charging demand point. The charging demand points are clustered to determine the charging demand marking points, the distance from each charging demand marking point to all distribution network nodes is calculated, and the distribution network node closest to the charging demand marking point is used as the alternative access node for the charging station.

4. The method for site selection and capacity determination of a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches according to claim 1, characterized in that: The multi-objective optimization model for the site selection of the photovoltaic storage and charging microgrid and the energy storage sizing described in step 4 aims to minimize the voltage offset, minimize the distribution network loss and minimize the energy storage investment cost, and sets the consistency of the remaining energy storage capacity at the beginning and end of the optimization cycle, the charging and discharging power limit of the energy storage, the charge state limit, the energy storage capacity limit, the node voltage limit, the real-time balance of the node power and the selection of the access location from the alternative nodes as constraints.

5. The method for site selection and capacity determination of a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches according to claim 1, characterized in that: The nonlinear weight ω calculation method used by the improved particle swarm algorithm in step 4 is: Where a and b are auxiliary parameters for adjusting the numerical range of inertia weight; K max is the maximum number of iterations; Z is the impact factor, which is calculated as follows: Where U ω is the flag bit, U ω =1 means that the population extreme value has not changed for 5 consecutive iterations, U ω = 0 means that the population extreme value has changed at least once in 5 consecutive iterations of the particle.

6. The method for site selection and capacity determination of a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches according to claim 1, characterized in that: The method for determining the feasible solution set in step 4 is: Set the number of particles to N p , each iteration will update N p The individual extreme value of particles is taken as the feasible solution after each iteration. max Iterations, finally generated K max ×N p feasible solutions, and all feasible solutions constitute the feasible solution set Ω k .

7. The method for site selection and capacity determination of a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches according to claim 1, characterized in that: The process of determining alternative solutions in step 4 is as follows: The feasible solution set Ω k Each feasible solution in is compared with all other feasible solutions one by one, and the feasible solutions with dominating relationships are deleted, and finally N feasible solutions that do not dominate each other are obtained. f Pareto optimal solutions are selected and used as alternatives.

8. The method for site selection and capacity determination of a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches according to claim 1, characterized in that: In step 4, the improved approach to ideal solution sorting method uses Manhattan distance to represent the distance between each alternative solution and the positive and negative ideal solutions.

9. A site selection and capacity determination system for a photovoltaic storage and charging microgrid connected to a distribution network with flexible interconnected branches based on the site selection and capacity determination method described in any one of claims 1 to 8, characterized in that Includes the following modules: Data acquisition module, used to obtain the original EV data and distribution network data required for site selection and capacity determination of photovoltaic storage and charging microgrid; The model building module uses the charging station site selection method based on EV charging demand prediction to determine the alternative access nodes of the charging station, use them as the alternative nodes for the photovoltaic storage and charging microgrid to access the distribution network, and determine the installable photovoltaic capacity on the alternative nodes; establish a multi-objective optimization model for the site selection and energy storage capacity determination of the photovoltaic storage and charging microgrid; The solution module uses the improved particle swarm algorithm to solve the multi-objective optimization model and obtain a feasible solution set. The Pareto optimal solution is selected from the feasible solution set and used as an alternative solution. The solution determination module uses the improved approximate ideal solution sorting method to determine the optimal solution for the site selection and energy storage capacity of the photovoltaic storage and charging microgrid from the alternative solutions, and supplements the determined charging station load and photovoltaic capacity configuration plan to obtain the optimal solution for the complete photovoltaic storage and charging microgrid to access the distribution network.

Images