Optimization method of electric vehicle power distribution network based on fuzzy Pareto optimization

Abstract

The invention relates to a fuzzy Pareto optimization-based electric vehicle power distribution network optimization method. The method comprises the following steps of S1, constructing a network load model of a distributed power generation (DG), a parallel capacitor (SC) and an electric vehicle (EV); s2, proposing a multi-objective function based on fuzzy Pareto optimization; s3, verifying the radial performance of a part needing network optimization in the power distribution network based on a graph theory; and S4, optimizing the network topology of the power distribution network based on fuzzy Pareto optimization. According to the invention, the system performance of the power distribution network can be improved, such as reducing power loss and improving voltage distribution.

Description

Technical Field

[0001] This application relates to the field of optimization of electric vehicle distribution networks, and particularly to an optimization method for an electric vehicle distribution network based on fuzzy Pareto optimization. Background Art

[0002] With the popularization of electric vehicles, the charging loads of electric vehicles in the distribution network are gradually increasing. It is necessary to manage the charging load technology efficiently and orderly so that users can enjoy high-quality electricity. In addition, the access of a large number of electric vehicle loads also brings some problems to the system performance of the power grid, such as power loss and voltage drop. At present, the distribution network can improve the power flow distribution of the power grid, increase the node voltage in the power system, and reduce the active power loss in the process of power transmission by means of distributed generation and shunt capacitors. Optimizing the network topology configuration of the distribution network can further reduce the negative impact of the access of electric vehicle loads to the distribution network. Summary of the Invention

[0003] The purpose of the embodiments of this application is to provide an optimization method for an electric vehicle distribution network based on fuzzy Pareto optimization, which reduces the active power loss in the distribution network, improves the voltage distribution to enhance the power grid system performance and provides users with high-quality electricity.

[0004] To achieve the above purpose, this application provides the following technical solutions:

[0005] The embodiments of this application provide an optimization method for an electric vehicle distribution network based on fuzzy Pareto optimization, including the following steps:

[0006] S1: Construct a network load model of distributed generation DG, shunt capacitor SC, and electric vehicle EV;

[0007] S2: Propose a multi-objective function based on fuzzy Pareto optimization;

[0008] S3: Verify the radiality of the part of the distribution network that needs network optimization based on graph theory;

[0009] S4: Optimize the network topology of the distribution network based on fuzzy Pareto optimization.

[0010] The active power and reactive power formulas of the constructed load model are:

[0011] P(i) = PL(i) - P DG (i) + P EV (i) (1)

[0012] Q(i) = QL(i) - Q SC (i) - Q DG (i) + Q EV (i) (2)

[0013] Where P(i) is the net active load, PL(i) is the active load of the i-th node in the distribution network system, P DG (i), P EV (i) are the DG capacity and the active power load of the electric vehicle charging station at the i-th node of the distribution system respectively. Correspondingly, Q(i) is the net reactive load, QL(i) is the reactive load of the i-th node of the distribution system, Q SC (i), Q DG (i), Q EV (i) are the reactive powers provided by the DG with lagging power factor, SC reactive injection and EV charging load at the i-th node of the distribution system respectively.

[0014] The equations for the active power and reactive power injected by DG are as follows:

[0015]

[0016] Where is the power angle of the DG unit.

[0017] The total load of the electric vehicle charging station at any time depends on the number of charging vehicles and the capacity of the charger to supply power to the electric vehicle battery. The active power load and reactive power load at the charging station are expressed as:

[0018] P EV (i) = NEV(i)P BL (4)

[0019]

[0020] Where NEV(i) is the number of electric vehicles at the i-th node, P BL is the maximum charging load of a single charger. is the power angle of the electric vehicle charger.

[0021] To achieve efficient power transportation, it is necessary to minimize the active power loss as much as possible. In the fuzzy algorithm, an objective function for minimizing the active power loss is proposed:

[0022]

[0023] Where Ploss Base is the basic active power loss at this iteration, Ploss BERC is the active power loss after successive iterative improvement of the algorithm.

[0024] J PLI is the ratio of the actual active power loss to the active power loss in the basic case. For this objective function J PLI , the improvement condition is:

[0025]

[0026] where K is the number of iterations in the algorithm;

[0027] The minimum node voltage improvement target forms a fuzzy objective function as follows:

[0028]

[0029] This fuzzy objective function represents μ V For values less than or equal to V MINB the value is taken as zero, and when the value is less than V L1 and greater than V MINB the value is assigned between 0 and 1, and for μ V the value is 1 when between V L1 and V L2 and 0 when greater than V L2 The values of V L1 and V L2 are taken as 1.0 and 1.05, and the minimum voltage improvement objective function is proposed:

[0030] J VMINB = μ V (9)

[0031] The second objective function J VMINB is to raise the minimum node voltage to between the standard limits. The minimum node voltage is modeled as a fuzzy Pareto objective function. If the minimum node voltage of the system is within a certain range, then J VMINB is 1, otherwise J VMINB is less than 1. Only when the objective function is improved is the configuration of the switch position updated. Therefore, the improvement condition for this objective function is:

[0032]

[0033] Adding electric vehicle loads in the distribution network will increase the line burden. The branch current should be less than the maximum branch current limit. The maximum branch current constraint objective function is proposed:

[0034]

[0035] where I i and I Ci are the branch current of the i-th branch and the rated current of the i-th branch respectively. When the maximum branch current ratio is less than 1, the value of J MBCR is taken as 1; when it exceeds 1, the value of J MBCR is taken as 0. The third objective function J MBCR is that the branch current must always be less than the maximum conductor current-carrying capacity. Only when J MBCRWhen the value is 1, the network configuration optimization is considered, and the improvement conditions of the objective function are as follows:

[0036]

[0037] Simultaneously minimize three objectives, J PLI 、J VMINB and J MBCR The fuzzy-based Pareto multi-objective vector J is represented as follows:

[0038]

[0039] where J is the Pareto multi-objective vector function, is the vector containing the open switch positions.

[0040] Specifically, the radiality of the part of the distribution network that needs network optimization verified based on graph theory is

[0041] Represent the distribution network as an undirected graph. The following formula is used to verify the radiality of the network:

[0042] TF = graphisspantree(G) (15)

[0043] where G is an N×N sparse matrix, the lower triangle of which represents the undirected graph of the distribution system, and the non-zero terms in matrix G indicate the existence of edges. If the network is radial, TF is 1; otherwise, TF is 0.

[0044] Specifically, the optimization of the distribution network topology based on fuzzy Pareto optimization is

[0045] Step (1): Read the line and load data of the distribution system;

[0046] Step (2): Read the positions and load data of DGs, SCs, and EV charging stations;

[0047] Step (3): Identify the first loop sequence;

[0048] Step (4): Let the starting iteration count K = 0, the loop sequence count LC = 0, and the branch count BC = 0;

[0049] Step (5): Increment the iteration count K = K + 1;

[0050] Step (6): Calculate the objective function J PLI ,J VMINB ,J MBCR ;

[0051] Step (7): Increment the loop sequence count LC = LC + 1;

[0052] Step (8): Close all branches of the loop sequence, except for branches shared with other loop sequences and branches that are in the open state due to the status of their respective loop sequences;

[0053] Step (9): Increment the branch count BC = BC + 1;

[0054] Step (10): Open branch "BC" in the loop, keeping the remaining branches in their original positions;

[0055] Step (11): If the radial property is satisfied, calculate the objective J PLI , J VMINB , J MBCR ;

[0056] Step (12): If the Pareto optimality condition is satisfied, then update J K PLI = J (BC,LC) PLI , J K VMINB = J (BC ,LC) VMINB , J K MBCR = J (BC,LC) MBCR = 1, and update the network configuration; otherwise, retain the previous network configuration and update the values of VMINB and Ploss base ;

[0057] Step (13): If BC < BC max lc , return to Step (9);

[0058] Step (14): If LC < LC max , return to Step (7);

[0059] Step (15): Calculate and

[0060] Step (16): If and proceed to the next step, otherwise, return to Step (5);

[0061] Step (17): Stop the iteration and store the optimal network configuration switch positions.

[0062] Compared with the prior art, the beneficial effects of the present invention are as follows: By constructing a network load model of distributed power generation units, shunt capacitors, and electric vehicle loads, and then from three aspects of reducing active power loss, improving voltage distribution, and restricting branch current, a multi-objective function based on fuzzy Pareto optimization is proposed. Only when one of all the objective functions is improved, the network configuration is updated. Then, radiality verification is carried out on the part that needs to be optimized to ensure the feasibility of the algorithm process. Finally, an algorithm process for optimizing the distribution network configuration based on fuzzy Pareto optimization is proposed, so as to achieve the goals of minimizing active power loss, improving voltage distribution, and restricting branch current, improving system performance, and enhancing the stability and reliability of the distribution network. Description of the Drawings

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0064] Figure 1 It is a flowchart of the method of the present invention.

[0065] Figure 2 It is a flowchart for optimizing the distribution network topology based on fuzzy Pareto optimization of the present invention. Detailed Embodiments

[0066] The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. It should be noted that: Similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0067] The term "comprising", "including" or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article or apparatus comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or apparatus. Without further limitation, an element defined by the phrase "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising the element.

[0068] The terms "first", "second", etc. are only used to distinguish one entity or operation from another entity or operation, and cannot be understood as indicating or implying relative importance, nor can they be understood as requiring or implying any actual relationship or order between these entities or operations.

[0069] The following is a description of the implementation mode of the present invention in conjunction with Figures 1 to 2 to introduce the implementation mode of the present invention, Figure 1 which is a flowchart of the optimization process of a distribution network based on a fuzzy Pareto optimization algorithm according to the present invention, specifically:

[0070] S01: Construct a network load model for distributed generation (DG), shunt capacitors (SC), and electric vehicles (EV).

[0071] S02: Propose a multi-objective function based on fuzzy Pareto optimization.

[0072] S03: Verify the radiality of the part of the distribution network that needs network optimization based on graph theory.

[0073] S04: Propose an algorithm flow for optimizing the network configuration of a distribution network based on fuzzy Pareto optimality.

[0074] In step S01, construct a network load model for distributed generation (DG), shunt capacitors (SC), and electric vehicles (EV).

[0075] First, construct a PQ load model for distributed generation, shunt capacitors, and electric vehicle charging stations for power flow analysis and to determine the initial performance of the target system for implementing this algorithm. In power flow analysis, distributed generation units are regarded as having positive active and negative reactive loads, and DG units are considered to operate with a power factor of 0.95 lag. Shunt capacitors (SC) are regarded as negative reactive loads. The active load and reactive load formulas for this load model are:

[0076] P(i) = PL(i) - P DG (i) + P EV (i) (1)

[0077] Q(i) = QL(i) - Q SC (i) - Q DG (i) + Q EV (i) (2)

[0078] Where P(i) is the net active load, and PL(i) is the active load of the i-th node in the distribution network system. P DG (i), P EV (i) are respectively the DG capacity of the i-th node in the distribution system and the active power load of the electric vehicle charging station. Correspondingly, Q(i) is the net reactive load, QL(i) is the reactive load of the i-th node in the distribution system, Q SC (i), Q DG (i), Q EV(i) The reactive power provided by the DG with lagging power factor, the reactive power injection of the SC, and the EV charging load for the i-th node of the distribution system respectively.

[0079] Since the DG is considered to operate at a lagging power factor of 0.95, the equations for the active power and reactive power injected by the DG are as follows:

[0080]

[0081] where is the power angle of the DG unit.

[0082] The total load of the electric vehicle charging station at any time depends on the number of charging vehicles and the capacity of the charger to supply power to the electric vehicle battery. The active power load and reactive power load at the charging station are expressed as:

[0083] P EV (i) = NEV(i)P BL (4)

[0084]

[0085] where NEV(i) is the number of electric vehicles at the i-th node, and P BL is the maximum charging load of a single charger. is the power angle of the electric vehicle charger. The electric vehicle charging load is only considered at the optimal charging station and is taken as zero at the remaining locations.

[0086] In step S02, a multi-objective function based on fuzzy Pareto optimization is proposed.

[0087] During the process of optimizing the distribution network configuration, reducing the active power loss, improving the voltage distribution, and branch current constraints are taken as the optimization objectives.

[0088] To achieve efficient power transmission, it is necessary to minimize the active power loss as much as possible. In the fuzzy algorithm, an objective function for minimizing the active power loss is proposed:

[0089]

[0090] where Ploss Base is the basic active power loss at this iteration. Ploss BERC is the active power loss after successive iterations and improvements of the algorithm.

[0091] J PLI is considered to be the ratio of the actual active power loss to the active power loss in the basic case. In this algorithm, when at least one of all objective functions is improved, the network configuration is considered for update. For this objective function J PLI , the improvement condition is:

[0092]

[0093] Where K is the number of iterations in the algorithm.

[0094] The voltage level of each node is a measure of the quality of reliable power supply to users, and the lowest node voltage should always be within the standard limit. The fuzzy algorithm is more suitable for defining the objective function as different groups. Therefore, the minimum node voltage improvement objective is formed as a fuzzy objective function as follows:

[0095]

[0096] This fuzzy objective function represents μ V For values less than or equal to V MINB are taken as zero. When this value is less than V L1 and greater than V MINB , this value is assigned between 0 and 1. μ V values are 1 between V L1 and V L2 , and 0 when greater than V L2 . V L1 and V L2 are taken as 1.0 and 1.05. The improved minimum voltage objective function is proposed:

[0097] J VMINB = μ V (9)

[0098] The second objective function J VMINB is to increase the minimum node voltage to between the standard limits. The minimum node voltage is modeled as a fuzzy Pareto objective function. If the minimum node voltage of the system is within a certain range, then J VMINB is 1, otherwise J VMINB is less than 1. Only when the objective function is improved is the configuration of the switch position updated. Therefore, the objective function improvement condition is:

[0099]

[0100] Increasing the electric vehicle load in the distribution network will increase the line burden. Therefore, it should be noted that the branch current should be less than the maximum branch current limit. The maximum branch current constraint objective function is proposed:

[0101]

[0102] Where I i and I Ci are the branch current of the i-th branch and the rated current of the i-th branch respectively. When the maximum branch current ratio is less than 1, the value of J MBCR is taken as 1; when it exceeds 1, JMBCR The value is taken as 0. The third objective function J MBCR is that the branch current must always be less than the maximum conductor current-carrying capacity. Only when J MBCR has a value of 1, is the optimization of the network configuration considered. Therefore, the improvement condition of this objective function is:

[0103]

[0104] vector does not perform worse than vector in the performance of any objective function, and is superior to vector in the performance of a certain objective function, then is said to be Pareto-dominant over Using the concept of Pareto dominance to explain Pareto optimization: Suppose there are N objective functions. For solution A, if no other solution can be found in the variable space that is Pareto-dominant over A, then solution A is a Pareto optimal solution.

[0105] Simultaneously minimizing the three objectives, J PLI 、J VMINB and J MBCR The fuzzy-based Pareto multi-objective vector J is represented as follows:

[0106]

[0107] where J is the Pareto multi-objective vector function, is the vector containing the positions of the open switches. In the distribution system, each tie line forms a loop. During normal operation, all tie lines will be in the open position. At any time, if the radial condition is satisfied. The total number of switches in the open position during the reconfiguration algorithm is equal to the total number of loops, and the objective function is calculated only when the radial condition is satisfied.

[0108] Step S03, verify the radiality of the part of the distribution network that needs network optimization based on graph theory.

[0109] For optimizing the network configuration, it is necessary to verify the radiality of the network before updating the solution. In network optimization, the algorithm calculates the multi-objective function by sequentially closing and opening switches, and the calculation of the multi-objective function is related to the vector of switch positions, which requires radial verification of the network in different states. A network is radial if each node is connected to the substation through a unique path without any loops to avoid unnecessary circulating currents. According to graph theory, a spanning tree is an undirected graph that contains all nodes and must not form a closed loop. If the undirected graph generated from the distribution network graph satisfies the conditions of the spanning tree during the reconstruction process, the distribution network is said to be radial. Breadth-first search and depth-first search are spanning tree verification algorithms based on graph theory.

[0110] The built-in functions in the MATLAB toolbox can be used to determine radiality. The radiality of the distribution network is verified by using the "graphisspantree" function available in MATLAB. This function determines whether the network is a spanning tree. The existence of a spanning tree verifies the radial property of the network. First, the distribution network is represented as an undirected graph. The following formula is used to verify the radiality of the network:

[0111] TF = graphisspantree(G) (15)

[0112] where G is an N×N sparse matrix, and its lower triangle represents the undirected graph of the distribution system. The non-zero terms in matrix G indicate the existence of edges. If the network is radial, TF is 1; otherwise, TF is 0.

[0113] S04: Propose an algorithmic process for the network topology optimization of a distribution network based on fuzzy Pareto optimality.

[0114] This paper presents a detailed discussion of the network reconfiguration algorithm based on Pareto optimality. First, the loop sequence is determined, and the optimization algorithm is executed according to the loop sequence. When the loop sequence of the reconfiguration operation is identified, the first iteration starts by closing the connection line of the first loop sequence. Subsequently, each branch of the loop sequence is opened, and the Pareto objective is calculated. Accordingly, the best combination of switch actions that result in reduced active power loss and improved voltage is updated. Only when at least one of the three objective functions J PLI , J VMINB , J MBCR is improved, the network configuration is improved and the Pareto solution is updated. Similarly, this process is repeated for all other loop sequences. Finally, if the difference between the current iteration objective function and the previous iteration objective function reaches a pre-specified minimum value, the iteration is terminated, and the optimal network configuration is stored. The following are the detailed algorithmic process steps based on fuzzy Pareto optimality.

[0115] Step (1): Read the distribution system line and load data.

[0116] Step (2): Read the DG, SC, and EV charging station location and load data.

[0117] Step (3): Identify the first loop sequence.

[0118] Step (4): Set the starting iteration count K = 0, loop sequence count LC = 0, and branch count BC = 0.

[0119] Step (5): Increment the iteration count K = K + 1.

[0120] Step (6): Calculate the objective function J through power flow analysis PLI , J VMINB , J MBCR .

[0121] Step (7): Increment the loop sequence count LC = LC + 1.

[0122] Step (8): Close all branches of this loop sequence, except for the branches shared with other loop sequences and the branches that are in the open state due to the status of their respective loop sequences.

[0123] Step (9): Increment the branch count BC = BC + 1.

[0124] Step (10): Open branch "BC" in the loop, keeping the remaining branches in their original positions.

[0125] Step (11): If the radial property is satisfied, calculate objective J through power flow analysis PLI , J VMINB , J MBCR .

[0126] Step (12): If the Pareto optimization condition is satisfied, then update J K PLI = J (BC,LC) PLI , J K VMINB = J (BC ,LC) VMINB , J K MBCR = J (BC,LC) MBCR = 1, and update the network configuration. Otherwise, retain the previous network configuration. Update the values of VMINB and Ploss base .

[0127] Step (13): If BC < BC maxlc , return to step (9).

[0128] Step (14): If LC < LC max , return to step (7).

[0129] Step (15): Calculate and

[0130] Step (16): If and proceed to the next step, otherwise, return to step (5).

[0131] Step (17): Stop the iteration and store the optimal network optimization configuration switch position.

[0132] The network load model of distributed generation, shunt capacitors, and electric vehicle loads proposed by the present invention starts from the load characteristics of distributed generation units, shunt capacitors, and electric vehicle loads, and is used for power flow analysis and calculation of the multi-objective function of this algorithm.

[0133] The multi-objective function based on fuzzy Pareto optimization proposed by the present invention aims to reduce active power loss, improve voltage distribution, and constrain branch current. A multi-objective function is proposed, and the concept of Pareto optimization is applied to stipulate the improvement conditions for the three objective functions.

[0134] The present invention proposes to verify the radiality of the part of the distribution network that needs network optimization based on graph theory. When the network satisfies the radial property, the objective function is calculated. For the radiality of the network, it can be verified through graph theory. According to graph theory, a spanning tree is an undirected graph that contains all nodes and does not form a closed loop. If the undirected graph generated by the distribution network meets the conditions of the spanning tree, the distribution network is said to have a radial property. It can also be verified using the built-in functions in matlab.

[0135] The flow of the distribution network optimization configuration algorithm based on fuzzy Pareto optimization proposed by the present invention focuses on improving the multi-objective function. It gradually verifies the radiality and calculates the objective function for different branches of different loop sequences. When the difference between the objective function of the current iteration and the objective function of the previous iteration reaches the set minimum value, the current network configuration is saved.

[0136] The above are only the embodiments of the present application and are not used to limit the protection scope of the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An optimization method for an electric vehicle distribution network based on fuzzy Pareto optimization, characterized in that, It includes the following steps: S1: Construct the network load models of distributed generation DG, shunt capacitor SC, and electric vehicle EV; S2: Propose a multi-objective function based on fuzzy Pareto optimization; S3: Verify the radiality of the part of the distribution network that needs network optimization based on graph theory; S4: Optimize the network topology of the distribution network based on fuzzy Pareto optimization.

2. The optimization method of an electric vehicle distribution network based on fuzzy Pareto optimization according to claim 1, characterized in that, The active load and reactive load formulas of the constructed load model are: P(i) = PL(i) - P DG (i) + P EV (i) (1) Q(i) = QL(i) - Q SC (i) - Q DG (i) + Q EV (i) (2) Where P(i) is the net active load, PL(i) is the active load of the i-th node in the distribution network system, P DG (i), P EV (i) are the DG capacity and the active power load of the electric vehicle charging station at the i-th node of the distribution system, respectively. Correspondingly, Q(i) is the net reactive load, QL(i) is the reactive load of the i-th node of the distribution system, Q SC (i), Q DG (i), Q EV (i) are the reactive powers provided by the DG with lagging power factor, the reactive power injection of the SC, and the EV charging load at the i-th node of the distribution system, respectively. The equations of the active power and reactive power injected by DG are as follows: wherein is the power angle of the DG unit, The total load of the electric vehicle charging station at any time depends on the number of charging vehicles and the capacity of the charger to supply power to the electric vehicle battery. The active power load and reactive power load at the charging station are expressed as: P EV (i) = NEV(i)P BL (4) where NEV(i) is the number of electric vehicles at the i-th node, and P BL is the maximum charging load of a single charger, is the power angle of the electric vehicle charger.

3. An optimization method for an electric vehicle distribution network based on fuzzy Pareto optimization according to claim 1, characterized in that To achieve efficient power transportation, it is necessary to minimize the active power loss as much as possible. In the fuzzy algorithm, an objective function for minimizing the active power loss is proposed: Among them, Ploss Base is the basic active power loss at this iteration number, and Ploss BERC is the active power loss after successive iterative improvement of the algorithm. J PLI is the ratio of the actual active power loss to the active power loss in the basic case, for this objective function J PLI , the improvement conditions are: where K is the number of iterations in the algorithm; The objective of improving the minimum node voltage forms a fuzzy objective function as follows: This fuzzy objective function represents μ V For values less than or equal to V MINB The value is taken as zero when the value is less than V L1 And greater than V MINB At this time, the value is assigned between 0 and 1, μ V The value of μ is 1 when it is between V L1 And V L2 And is 0 when greater than V L2 V L1 And V L2 The values of V are taken as 1.0 and 1.05, and an improved minimum voltage objective function is proposed: J VMINB = μ V (9) The second objective function J VMINB is to raise the minimum nodal voltage to the standard limit value. The minimum nodal voltage is modeled as a fuzzy Pareto objective function. If the minimum nodal voltage of the system is within a certain range, then J VMINB is 1; otherwise, J VMINB is less than 1. Only when the objective function is improved will the configuration of the switch positions be updated. Therefore, the improvement condition for this objective function is as follows: Increasing the electric vehicle load in the distribution network will increase the line burden. The branch current should be less than the maximum branch current limit. An objective function for the maximum branch current constraint is proposed: where I i and I Ci are the branch current of the i-th branch and the rated current of the i-th branch respectively. When the maximum branch current ratio is less than 1, the value of J MBCR is taken as 1; When it exceeds 1, J MBCR takes the value of 0, and the third objective function J MBCR is that the branch current must always be less than the maximum conductor current-carrying capacity. Only when the value of J MBCR is 1, the optimization of the network configuration is considered. The improvement conditions of this objective function are: Simultaneously minimize three objectives, J PLI , J VMINB and J MBCR The fuzzy-based Pareto multi-objective vector J representation for J, J, and J is as follows: where J is the Pareto multi-objective vector function, is a vector containing the positions of the disconnecting switches.

4. An optimization method for an electric vehicle distribution network based on fuzzy Pareto optimization according to claim 1, characterized in that, The verification of the radiality of the part of the distribution network that needs network optimization based on graph theory is specifically as follows: The distribution network is represented as an undirected graph. The following formula is used to verify the radiality of the network: TF = graphisspantree(G) (15) where G is an N×N sparse matrix, and its lower triangle represents the undirected graph of the distribution system. The non-zero terms in matrix G indicate the existence of edges. If the network is radial, TF is 1; otherwise, TF is 0.

5. An optimization method for an electric vehicle distribution network based on fuzzy Pareto optimization according to claim 1, characterized in that, The optimization of the distribution network topology based on fuzzy Pareto optimization is specifically as follows: Step (1): Read the line and load data of the distribution system; Step (2): Read the location and load data of DG, SC, and EV charging stations; Step (3): Identify the first loop sequence; Step (4): Let the starting iteration count K = 0, the loop sequence count LC = 0, and the branch count BC = 0; Step (5): Increment the iteration count K = K + 1; Step (6): Calculate the objective function J through power flow analysis PLI , J VMINB , J MBCR ; Step (7): Increment the loop sequence count LC = LC + 1; Step (8): Close all branches of this loop sequence, except for the branches shared with other loop sequences and the branches that are in the disconnected state due to the state of their respective loop sequences; Step (9): Increment the branch count BC = BC + 1; Step (10): Open the branch "BC" in the loop, keeping the remaining branches in their original positions; Step (11): If the radial property is satisfied, calculate the target J through power flow analysis PLI , J VMINB , J MBCR ; Step (12): If the Pareto optimality condition is satisfied, update J K PLI = J (BC,LC) PLI , J K VMINB = J (BC,LC) VMINB , J K MBCR = J (BC,LC) MBCR = 1, and update the network configuration; otherwise, retain the previous network configuration and update the values of VMINB and Ploss base ; Step (13): If BC < BC max lc , return to Step (9); Step (14): If LC < LC max , return to step (7); Step (15): Calculate and Step (16): If and proceed to the next step; otherwise, return to step (5). Step (17): Stop the iteration and store the switch positions of the optimal network optimization configuration.

Images