Electric vehicle charging station site selection optimization method, system and equipment

Abstract

This application relates to a method, system, and device for optimizing the site selection of electric vehicle charging stations, and relates to the technical field of charging station planning. The method comprises: constructing an OD matrix for a target area based on multiple nodes of charging station site selection, and performing real-time updates to obtain an OD matrix sequence; analyzing the flow coefficient and flow fluctuation coefficient based on the OD matrix sequence, and configuring flow weights and load weights; and optimizing the site selection within multiple nodes based on the flow weights and load weights based on particle swarm optimization, obtaining multiple optimal nodes as the site selection optimization results. The present invention solves the problem that traditional charging station site selection fails to fully consider the dynamic changes in traffic flow and the impact of power grid load, resulting in insufficient accuracy of site selection plans.

Description

Technical Field

[0001] The present application relates to the field of charging station planning, and in particular to a method, system, and device for optimizing the site selection of electric vehicle charging stations. Background Art

[0002] With the rapid development of electric vehicle charging infrastructure, the scientific site selection of charging stations has become a key step in improving the efficiency of charging resource utilization and the stability of power grid operations. Currently, traditional charging station site selection techniques mostly use static planning methods that simply consider single factors such as traffic flow or grid load. This leads to problems such as weak dynamic adaptability and insufficient site selection accuracy.

[0003] Existing site selection methods are optimized only based on fixed traffic flow data and grid load thresholds, resulting in significant deviations between site selection plans and actual charging demand and grid carrying capacity. This not only causes waste of charging resources and grid operation risks, but also fails to adapt to the requirements of the intelligent development of electric vehicle charging networks for the accuracy and reliability of site selection plans. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a method, system and equipment for optimizing the site selection of electric vehicle charging stations, which improves the charging station site selection's ability to accurately respond to dynamic changes in traffic flow and grid load shocks, and solves the problem that the existing technology does not fully integrate multi-dimensional needs, resulting in insufficient site selection accuracy.

[0005] The embodiments of this application disclose the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a method for optimizing the site selection of an electric vehicle charging station, the method comprising:

[0007] Based on multiple nodes of charging station site selection, the OD matrix of the target area is constructed and updated in real time to obtain the OD matrix sequence;

[0008] According to the OD matrix sequence, the flow coefficient and the flow fluctuation coefficient are analyzed, and the flow weight and the load weight are configured;

[0009] Based on particle swarm optimization, location optimization is performed within multiple nodes according to the traffic weight and load weight, and multiple optimal nodes are obtained as location optimization results, wherein the optimization step size is configured according to the traffic weight and load weight for optimization.

[0010] In a second aspect, an embodiment of the present application provides a system for optimizing the site selection of electric vehicle charging stations, the system comprising:

[0011] The OD matrix construction and update module is used to construct the OD matrix of the target area based on multiple nodes of the charging station site selection, and update it in real time to obtain the OD matrix sequence;

[0012] A flow load weight configuration module, configured to analyze the flow coefficient and the flow fluctuation coefficient according to the OD matrix sequence, and configure the flow weight and the load weight;

[0013] The weight-driven site selection optimization module is used to perform site selection optimization within multiple nodes based on the flow weight and load weight based on particle swarm optimization, and obtain multiple optimal nodes as site selection optimization results, wherein the optimization step size is configured according to the flow weight and load weight for optimization.

[0014] In a third aspect, an embodiment of the present application provides a computer device, comprising: a memory for storing a computer program; and a processor for reading and executing the computer program, thereby implementing the electric vehicle charging station site selection optimization method of the first aspect.

[0015] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0016] This application proposes a method, system and equipment for optimizing the site selection of electric vehicle charging stations, which realizes the precise site selection of charging stations through multi-dimensional traffic data fusion and dynamic weight optimization. First, the OD matrix of the target area is constructed and updated in real time to form a matrix sequence that reflects the dynamic changes in traffic flow; then the matrix data is analyzed to obtain the flow coefficient and fluctuation coefficient, and the modified weight is generated by combining the preset weights, and the final flow weight and load weight are obtained through matching. Further based on the particle swarm optimization algorithm, the initial node is randomly selected, the traffic frequency set is extracted to calculate the traffic rate, fluctuation rate and site fitness, the optimization step size is dynamically configured through weighted calculation, the node position is iteratively updated, and finally the optimal site node is obtained by convergence. This method solves the problem of deviation caused by traditional site selection relying on static data and single factors. Through dynamic weights and particle swarm optimization, the applicability of charging station layout and grid load balancing in complex traffic scenarios are improved, providing an intelligent solution for charging network planning.

[0017] The technical solution of this application solves the site selection deviation problem caused by the traditional method of simply considering traffic flow or grid load as a single factor by integrating historical traffic flow data, real-time fluctuation characteristics and dynamic weight parameters. It effectively improves the comprehensive applicability of charging station layout and grid load balance in complex traffic scenarios, and provides a reliable solution for the scientific site selection of electric vehicle charging stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 A flow chart of a method for optimizing the site selection of an electric vehicle charging station provided in an embodiment of the present application;

[0020] Figure 2 A schematic diagram of the structure of an electric vehicle charging station site selection optimization system provided in an embodiment of the present application;

[0021] Figure 3 A schematic diagram of the structure of a computer device provided in an embodiment of the present application.

[0022] In the accompanying drawings, the components represented by the reference numerals are described as follows:

[0023] OD matrix construction and updating module 01, traffic load weight configuration module 02, weight-driven location optimization module 03, computer device 300, memory 310, processor 320, computer program 311. DETAILED DESCRIPTION

[0024] The present application provides a method, system and device for optimizing the site selection of electric vehicle charging stations, which are used to solve the technical problems existing in the prior art, such as static planning leading to mismatch between supply and demand, failure to combine the dynamics of traffic flow and the multi-dimensional demands of grid load impact, and insufficient site selection accuracy.

[0025] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0026] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the described features. In the description of this application, "plurality" means two or more, unless otherwise specifically specified.

[0027] In the description of this application, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art will recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in this application.

[0028] Example 1, as shown in the attached Figure 1 As shown, the present application provides a method for optimizing the site selection of electric vehicle charging stations, the method comprising the following steps:

[0029] S110: Based on multiple nodes of charging station site selection, construct an OD matrix of the target area and update it in real time to obtain an OD matrix sequence;

[0030] In the embodiment of the present application, in the scenario of site selection optimization for electric vehicle charging stations, in order to accurately analyze the changes in traffic flow at different times and locations, it is necessary to collect dynamic traffic data based on the site selection nodes and construct a matrix sequence.

[0031] Specifically, for multiple nodes such as shopping malls and office buildings where charging stations are to be located, traffic monitoring equipment is used to collect traffic data between nodes in real time within a preset time range, including information such as traffic direction and number of passes.

[0032] Furthermore, the collected inter-node traffic data are processed to extract the number of traffic trips between every two nodes, and an OD matrix is ​​constructed based on multiple sets of node traffic data.

[0033] Furthermore, a dynamic update method is adopted to repeat the above data collection and matrix construction process at preset time intervals to form a continuous OD matrix sequence to reflect the real-time change trend of traffic flow.

[0034] This step provides accurate traffic flow conditions at different times and locations for charging station site selection by collecting data in real time and updating the matrix regularly, providing a data basis for subsequent optimization of electric vehicle charging station site selection.

[0035] Step S110 of the method provided in the embodiment of the present application includes:

[0036] Obtain multiple nodes for charging station site selection;

[0037] Based on traffic monitoring data, collect traffic data between multiple nodes within a preset time range;

[0038] Construct an OD matrix based on the traffic data between multiple nodes;

[0039] OD matrices within multiple preset time ranges are updated and constructed in real time to obtain OD matrix sequences.

[0040] In the embodiment of the present application, in order to accurately capture the distribution and variation characteristics of traffic flow in time and space, it is necessary to collect dynamic traffic data according to the site selection nodes and construct an OD matrix sequence.

[0041] Specifically, first, based on the layout of urban functional areas and charging demand forecasts, the nodes for charging station locations are determined, such as shopping malls, office buildings, parking lots and other places that have the conditions for building charging piles.

[0042] Among them, the first type of nodes are large shopping malls in urban commercial centers, which have large daily traffic volume and long parking time, and are suitable for the layout of fast charging stations; the second type of nodes are office building parking lots, which have concentrated traffic volume during peak hours in the morning and evening on weekdays and can be configured with an appropriate number of charging piles; the third type of nodes are public parking lots, which cover the surrounding residential areas and meet daily charging needs.

[0043] Furthermore, based on the traffic data between each node, the OD matrix is ​​constructed

[0044] The step of “constructing an OD matrix based on traffic data between multiple nodes” in the method provided in the embodiment of the present application includes:

[0045] According to the traffic data between multiple nodes, the number of traffic between every two nodes is extracted;

[0046] Based on the number of traffic flows between multiple groups of nodes, an OD matrix is ​​constructed, where each group of nodes includes two nodes.

[0047] In the embodiment of the present application, when constructing the OD matrix, the collected traffic data must first be structured to form a matrix model that reflects the traffic flow characteristics between nodes.

[0048] Specifically, after determining multiple nodes to be sited, firstly, the traffic data between each node within a preset time range is collected in real time through traffic monitoring equipment (such as road checkpoint cameras, RFID sensors, geomagnetic detectors, etc.).

[0049] Taking shopping mall node A, office building node B and parking lot node C as examples, information such as the number of vehicles from A to B, the number of vehicles from B to A, the number of vehicles from A to C, the number of vehicles from C to A, the number of vehicles from B to C, and the number of vehicles from C to B are collected. This information includes traffic direction, number of passes, etc.

[0050] For example, during the morning rush hour (7:00-9:00), 120 vehicles pass from node A to node B, 85 vehicles pass from node B to node A, 50 vehicles pass from node A to node C, 30 vehicles pass from node C to node A, 40 vehicles pass from node B to node C, and 60 vehicles pass from node C to node B.

[0051] Furthermore, the collected traffic data between the nodes is processed to extract the number of traffic flows between every two nodes.

[0052] For example, the number of vehicles traveling from node A to node B is considered one travel count between the two nodes, and the number of vehicles traveling from node B to node A is considered another travel count, and so on, to obtain multiple sets of travel counts between nodes. If there are N location nodes, then a total of N × (N-1) sets of inter-node travel count data are generated.

[0053] Furthermore, an OD matrix is ​​constructed based on the number of traffic flows between multiple groups of nodes. The rows of the matrix represent the starting nodes, i.e., the traffic generating ends, and the columns represent the ending nodes, i.e., the traffic attracting ends. The matrix element values ​​are the number of traffic flows from the corresponding starting point to the ending point.

[0054] For example, a 3×3 OD matrix is ​​constructed by taking the three nodes A, B, and C as examples, where A to A, B to B, and C to C are usually 0 or a very small value, and the remaining elements are filled with the number of trips between the corresponding nodes, such as 120 from A to B, 85 from B to A, etc.; if a new residential node D is added and the number of nodes is expanded to 4, the dimension of the OD matrix becomes 4×4, and the data on the number of trips between nodes such as A and D, B and D, and C and D need to be supplemented to ensure that the matrix fully reflects the traffic flow distribution between all nodes.

[0055] This OD matrix quantifies the direction and volume of traffic flow between nodes, providing fundamental data support for subsequent optimization of EV charging station locations. For example, a high number of trips from A to B in the matrix can reflect the high commuting demand from a shopping mall to an office building, thus guiding the placement of charging stations near node B to match actual traffic demand.

[0056] Furthermore, to ensure that the OD matrix can dynamically reflect the real-time changes in traffic flow, it needs to be updated periodically to construct a matrix sequence.

[0057] Specifically, by periodically updating, that is, repeating the traffic data collection, inter-node traffic frequency extraction and OD matrix construction at fixed time intervals, a continuous matrix sequence is formed.

[0058] For example, based on the first set of OD matrices constructed during the morning rush hour of 7:00-8:00, the traffic data between the shopping mall node A, the office building node B, and the parking lot node C are re-collected during the period of 8:00-9:00. For example, the number of passes from A to B during this period is updated to 150 vehicles and from B to A to 100 vehicles, and the traffic between the remaining nodes is refreshed synchronously.

[0059] Furthermore, after constructing the second set of OD matrices based on the new data, a time series is formed with the first set of matrices, and so on, generating a set of matrices every hour.

[0060] At the same time, multiple sets of OD matrices are stored and managed in chronological order through the cloud server to ensure the temporal continuity of the matrix sequence.

[0061] In addition, if a residential node D is added during a certain period, the matrix dimension will be automatically expanded to 4×4, and traffic data between D and other nodes (such as A to D, D to B, etc.) will be synchronously collected in subsequent update cycles, so that the matrix sequence can adapt to the increase in nodes.

[0062] Finally, the OD matrix sequence obtained through the above steps records the dynamic changes in traffic flow between nodes in different time periods, providing time-series data support for analyzing traffic flow changes and charging station site selection.

[0063] For example, by comparing the number of trips from A to B in the matrix sequence from 7:00 to 9:00, which increases from 120 vehicles to 150 vehicles, it can be determined that the commuting flow from the shopping mall to the office building in this time zone is on the rise, thereby guiding the dynamic adjustment of the charging station site selection plan according to the real-time traffic characteristics.

[0064] S120: Analyze the flow coefficient and flow fluctuation coefficient according to the OD matrix sequence, and configure the flow weight and load weight;

[0065] In the embodiment of the present application, in the scenario of site selection optimization for electric vehicle charging stations, in order to accurately quantify the influence weight of traffic flow characteristics on site selection, it is necessary to carry out multi-dimensional traffic analysis based on the OD matrix sequence.

[0066] Specifically, for the constructed OD matrix sequence, the average number of traffic times among all nodes is calculated first to reflect the overall traffic flow size in the target area.

[0067] Furthermore, the calculated average traffic number is compared with the preset traffic number to obtain a flow coefficient, which is multiplied by the preset flow weight to obtain a modified flow weight.

[0068] Furthermore, the discrete parameters of the traffic frequency in the OD matrix sequence are calculated to reflect the discrete degree of the overall traffic frequency.

[0069] Furthermore, the traffic frequency discrete parameter is ratioed with the preset traffic frequency discrete parameter to obtain a flow fluctuation coefficient, and the coefficient is multiplied by the preset load weight to obtain a modified load weight.

[0070] Finally, based on the corrected flow weight and load weight, the flow weight and load weight used for charging station site selection are obtained through comprehensive calculation, providing a quantitative basis for subsequent site selection optimization.

[0071] This step accurately quantifies the characteristics of traffic flow by analyzing the flow coefficient and flow fluctuation coefficient, providing a scientific basis for the reasonable configuration of flow weight and load weight.

[0072] Step S120 in the method provided in the embodiment of the present application includes:

[0073] According to the OD matrix sequence, the average traffic frequency is calculated;

[0074] Calculating a ratio of the average traffic number to a preset traffic number as a flow coefficient, and multiplying the ratio by a preset flow weight to obtain a modified flow weight;

[0075] Calculating the traffic frequency discrete parameter of the OD matrix sequence;

[0076] Calculating a ratio of the traffic frequency discrete parameter to a preset traffic frequency discrete parameter as a flow fluctuation coefficient, and multiplying the ratio by a preset load weight to obtain a modified load weight;

[0077] The flow weight and the load weight are calculated based on the modified flow weight and the modified load weight.

[0078] In the embodiment of the present application, in order to accurately distinguish the influence weight of traffic flow characteristics on the location of charging stations, it is necessary to perform a multi-dimensional weight analysis based on the OD matrix sequence.

[0079] Specifically, based on the constructed OD matrix sequence, the number of traffic flows between all nodes is first summarized and the average is calculated to obtain the average number of traffic flows, thereby reflecting the scale of the overall traffic flow in the target area.

[0080] For example, the OD matrix sequence of a certain area covers matrices for the morning, afternoon, and evening time periods. The number of trips from node A to B is 120, 80, and 100, respectively, and the number of trips from node B to C is 50, 30, and 40, respectively. The number of trips between all nodes is summed up and the average is taken ((120+80+100+50+30+40) / 6=420 / 6=70 times), which means the average number of trips in the area is 70.

[0081] Furthermore, the calculated average traffic count is compared to the preset traffic count to obtain a flow coefficient, which reflects the degree of deviation between the actual traffic flow in the target area and the baseline flow. For example, a flow coefficient greater than 1 indicates that the actual flow is higher than the baseline level, while a coefficient greater than 1 indicates that the actual flow is lower than the baseline level.

[0082] For example, if the preset number of trips is 100 and the average number of trips in a certain area is calculated to be 70, the flow coefficient is 70 / 100 = 0.7. This flow coefficient indicates that the actual traffic flow in the target area is 30% lower than the preset baseline value, reflecting that the current traffic demand in the area is relatively weak.

[0083] Furthermore, the obtained flow coefficient is multiplied by the preset flow weight to obtain a corrected flow weight.

[0084] Among them, the preset traffic frequency values ​​and preset flow weights are determined based on historical traffic statistics, urban planning standards or industry benchmark values, ensuring that the preset values ​​are scientific and referenceable, so that the flow coefficient can accurately reflect the difference between actual traffic characteristics and benchmark conditions.

[0085] For example, if the preset traffic weight is 40%, the corrected traffic weight is 0.7×40%=28%. That is, the traffic weight ratio needs to be adjusted according to the actual traffic attenuation so that the charging station site selection decision can better meet the actual needs of regional traffic.

[0086] At the same time, in order to further analyze the fluctuation of traffic flow, it is necessary to calculate the discrete parameters of the traffic frequency in the OD matrix sequence (that is, to obtain the variance of all the times) to reflect the degree of discreteness of the overall traffic frequency changes.

[0087] For example, in the OD matrix sequence of a certain region, the number of trips from node A to B is 100, 150, and 110, respectively, and the number of trips from node B to C is 40, 60, and 50, respectively. The total number of trips is summed up to 100, 150, 110, 40, 60, and 50, with an average of (100+150+110+40+60+50) / 6=90. The variance is calculated as [(100-90)²+(150-90)²+(110-90)²+(40-90)²+(60-90)²+(50-90)²] / 6≈9100 / 6≈1516.67, meaning that the discrete parameter of the number of trips in the OD matrix sequence of this region is 1516.67.

[0088] Furthermore, based on the comparison and evaluation of the obtained discrete parameter of traffic frequency with the preset discrete parameter of traffic frequency, a flow fluctuation coefficient is obtained, which is used to reflect the deviation of the fluctuation degree of traffic flow in the target area from the benchmark fluctuation level.

[0089] Among them, when the flow fluctuation coefficient is greater than 1, it indicates that the actual flow fluctuation level is higher than the benchmark level, otherwise it is lower than the benchmark level.

[0090] For example, if the preset traffic frequency discrete parameter is 1300 and the calculated traffic frequency discrete parameter for a certain area is 1516.67, then the traffic flow fluctuation coefficient is 1516.67 / 1300 ≈ 1.17. This traffic flow fluctuation coefficient indicates that the degree of traffic flow fluctuation in the target area is 17% higher than the preset baseline value, reflecting the relatively poor traffic flow stability in this area.

[0091] Furthermore, the obtained flow fluctuation coefficient is multiplied by the preset load weight to obtain a corrected load weight.

[0092] Similarly, the preset traffic frequency discrete parameters and the preset load weights are determined based on the fluctuation characteristics of historical traffic data, urban traffic management regulations or industry load balancing standards, so as to ensure that the flow fluctuation coefficient can accurately reflect the difference between the actual flow fluctuation characteristics and the standard state.

[0093] For example, if the preset load weight is 30%, the corrected load weight is 1.17×30%=35.1%. That is, the load weight ratio needs to be increased according to the actual traffic fluctuations so that the charging station site selection decision can better adapt to the fluctuating characteristics of regional traffic flow.

[0094] Furthermore, according to the obtained corrected flow weight and corrected load weight, the final flow weight and load weight are obtained through calculation.

[0095] In the method provided in the embodiment of the present application, the step of “calculating the flow weight and the load weight according to the modified flow weight and the modified load weight” includes:

[0096] Calculating the ratio of the modified flow weight to the sum of the modified flow weight and the modified load weight to obtain a flow weight;

[0097] The ratio of the corrected load weight to the sum of the corrected flow weight and the corrected load weight is calculated to obtain the load weight.

[0098] In the embodiment of the present application, in order to convert the corrected traffic weight and load weight into decision indicators that can directly guide the site selection of charging stations, it is necessary to achieve a scientific distribution of the two through weight ratio calculation.

[0099] Specifically, the corrected traffic weight and the corrected load weight are first summed to obtain the total weight, and then the proportion of the two corrected weights in the total is calculated respectively to obtain the final traffic weight and load weight, that is, "traffic weight = corrected traffic weight / the sum of corrected traffic weight and corrected load weight" and "load weight = corrected load weight / the sum of corrected traffic weight and corrected load weight".

[0100] For example, if the calculated corrected traffic weight for a region is 28% and the corrected load weight is 35.1%, the total weight is 28% + 35.1% = 63.1%. In this case, the traffic weight is 28% / 63.1% ≈ 44.4%, and the load weight is 35.1% / 63.1% ≈ 55.6%.

[0101] Among them, this calculation method ensures that the sum of the two weights is 1, so that the weight index can fully reflect the impact of traffic flow scale and fluctuation characteristics on site selection.

[0102] This step eliminates absolute numerical differences in the modified weights through weighted ratio calculations, making the flow weights and load weights relatively comparable and more instructive for decision-making. For example, when the modified flow weight is higher than the modified load weight, the final flow weight accounts for a larger proportion, indicating that site selection decisions should prioritize traffic flow scale; conversely, when the modified load weight is lower, the focus should be on load balancing requirements caused by traffic fluctuations.

[0103] S130: Based on particle swarm optimization, perform site selection optimization within multiple nodes according to the traffic weight and load weight, and obtain multiple optimal nodes as site selection optimization results, wherein the optimization step size is configured according to the traffic weight and load weight for optimization.

[0104] In the embodiment of the present application, in the scenario of site selection optimization for electric vehicle charging stations, in order to improve the adaptability of the site selection plan to the actual traffic flow characteristics, it is necessary to combine the flow weight and load weight and use the particle swarm optimization algorithm to achieve iterative optimization of the node position.

[0105] Specifically, based on the particle swarm optimization mechanism, according to the preset number of nodes M, M first nodes are randomly selected from the candidate node set as the initial site selection nodes, and each node corresponds to a particle in the particle swarm.

[0106] Furthermore, based on the OD matrix sequence, traffic data for M first nodes is extracted to form M sets of first-node traffic times. Taking each first node as the starting point or end point, the traffic times between it and other nodes are summarized to calculate M first traffic rates and M first traffic fluctuations.

[0107] Furthermore, combining the obtained flow weights and load weights, the first location fitness is calculated using the formula: "First location fitness = flow weight × first traffic rate + load weight × (1 − first traffic volatility)." This formula transforms the scale and volatility characteristics of traffic flow into a location fitness indicator through weighted ratios.

[0108] Furthermore, M optimization step sizes are configured according to the M first traffic rates and the M first traffic fluctuation rates in combination with the flow weight and the load weight.

[0109] Furthermore, the M first nodes are updated according to M optimization steps to obtain M second nodes, and through an iterative optimization process, such as setting the maximum number of iterations or the fitness convergence threshold, the node positions are continuously adjusted and the site fitness is calculated until multiple optimal nodes with the largest site fitness are obtained after convergence as the final site optimization result.

[0110] This step achieves a dynamic response of the optimization step to traffic flow characteristics through the dual effects of flow weight and load weight. That is, the higher the flow weight, the more the optimization process tends to select high-flow areas, and the higher the load weight, the more it tends to avoid areas with drastic flow fluctuations. This ensures that the final site selection plan can both meet charging needs and balance load pressure, effectively improving the accuracy of site selection optimization.

[0111] Step S130 in the method provided in the embodiment of the present application includes:

[0112] Based on particle swarm optimization, according to the preset number of nodes M, M first nodes are randomly selected as site selection nodes;

[0113] According to the OD matrix sequence, extracting M first node traffic frequency sets of M first nodes, calculating M first traffic rates and M first traffic volatility rates, and calculating first location fitness;

[0114] configuring M optimization step lengths according to the M first traffic rates and the M first traffic fluctuation rates, wherein the optimization step length includes the number of node movements;

[0115] Use M optimization steps to update M first nodes and obtain M second nodes;

[0116] Perform iterative optimization of the site selection nodes, and after convergence, obtain multiple optimal nodes with the largest site selection fitness as the site selection optimization results.

[0117] In the embodiment of the present application, in order to achieve accurate optimization of the site selection of electric vehicle charging stations, it is necessary to combine the flow weight and load weight and use the particle swarm optimization algorithm to dynamically adjust the site selection node position.

[0118] Specifically, based on the particle swarm optimization mechanism, according to the pre-set number of nodes M, M first nodes are randomly selected from all candidate nodes in the target area as the initial site selection nodes. Each node corresponds to a particle in the particle swarm, laying the foundation for subsequent iterative optimization.

[0119] Furthermore, based on the constructed OD matrix sequence, traffic data of M first nodes are extracted to form M first node traffic frequency sets, and M first traffic rates, M first traffic fluctuation rates and first location fitness are calculated.

[0120] In the method provided in the embodiment of the present application, the step of “extracting M first-node traffic frequency sets of M first nodes according to the OD matrix sequence, calculating M first traffic rates and M first traffic volatility rates, and calculating first site fitness” includes:

[0121] Extracting M first-node traffic frequency sets of M first nodes according to the OD matrix sequence, wherein each first-node traffic frequency set includes multiple traffic times of the first node in multiple OD matrices;

[0122] Calculating the ratio of the total traffic frequency of each first node traffic frequency set to the sum of the traffic frequencies in the OD matrix sequence to obtain M first traffic rates;

[0123] Calculate the maximum deviation between the minimum and maximum traffic times in the M first-node traffic times set respectively, and obtain M first traffic fluctuation rates;

[0124] The flow weight and load weight are used to perform weighted calculation on the difference between the M first traffic rates and 1 minus the M first traffic fluctuation rates, respectively, to obtain M first node location fitnesses, and the first location fitnesses are obtained by calculating the average.

[0125] In the embodiment of the present application, in order to accurately evaluate the suitability of each site selection node for the layout of electric vehicle charging stations, it is necessary to quantitatively analyze the traffic flow scale and fluctuation characteristics of each node based on the OD matrix sequence.

[0126] Specifically, after the initial M first nodes are selected, the traffic data of each node is extracted from the constructed OD matrix sequence to form a corresponding first node traffic frequency set.

[0127] For example, assume M = 3, select shopping mall A, office building B, and residential area C as the first node, and the OD matrix sequence contains matrices for the morning, afternoon, and evening time periods. In this case, for shopping mall A, the number of trips with all other nodes (e.g., A to B, B to A, A to C, and C to A) during these three time periods are extracted from the OD matrix sequence to form the first node trip count set for shopping mall A. Similarly, the number of trips with office building B and residential area C is extracted.

[0128] Furthermore, each first node traffic frequency set is processed to calculate the first traffic rate of each node, which is obtained by the formula "first traffic rate = total traffic frequency in the first node traffic frequency set / total traffic frequency in the OD matrix sequence".

[0129] For example, if the total number of traffic times in the first node traffic count set for shopping mall A is 300 (120 during the morning rush hour, 80 during the afternoon rush hour, and 100 during the evening rush hour), and the total number of traffic times in the entire OD matrix sequence is 1000, then the first traffic rate for shopping mall A is 300 / 1000 = 0.3. Using the same calculation method, the first traffic rates for M first nodes, such as office building B and residential area C, can be obtained.

[0130] Furthermore, in order to measure the fluctuation degree of the traffic flow of each node, the maximum deviation amplitude of the traffic times in each first node traffic time set is calculated to obtain the first traffic fluctuation rate.

[0131] For example, within the set of traffic counts for the first node of Office Building B, the minimum traffic count is 30 (during the lunch hour) and the maximum is 60 (during the evening rush hour). Therefore, the maximum deviation is (60-30) / 60 = 0.5, indicating that the first traffic volatility of Office Building B is 0.5. This method is used to calculate the first traffic volatility for all M first nodes.

[0132] Furthermore, based on the obtained flow weights and load weights, a weighted calculation is performed on the difference between the M first traffic rates and 1 minus the M first traffic volatility rates to obtain the M first node location fitnesses, which are obtained using the formula "first location fitness = flow weight × first traffic rate + load weight × (1 − first traffic volatility)".

[0133] For example, if the traffic weight is 40%, the load weight is 60%, the first traffic rate of mall A is 0.3, and the first traffic volatility is 0.4, then the location fitness of the first node of mall A is 40% × 0.3 + 60% × (1 - 0.4) = 0.48. The same calculation process can be used to obtain the location fitness of each of the M first nodes.

[0134] Finally, the M first-node location fitnesses are averaged to obtain the first location fitness (i.e., first-node location fitness = first-node location fitness / number of nodes M). For example, if the first-node location fitnesses of nodes A, B, and C are 0.48, 0.52, and 0.45, respectively, then the first location fitness is (0.48+0.52+0.45) / 3≈0.483.

[0135] Among them, the higher the first site fitness, the better the comprehensive performance of the currently selected M first nodes in meeting charging needs and balancing grid load, which will be more helpful to achieve a more reasonable electric vehicle charging station site selection plan.

[0136] In the method provided in the embodiment of the present application, the step of “configuring M optimization step sizes according to the M first traffic rates and the M first traffic fluctuation rates” includes:

[0137] Calculating a ratio of each first traffic rate to an average of the M first traffic rates to obtain M first flow adjustment coefficients;

[0138] Calculating a ratio of each first traffic fluctuation rate to an average of the M first traffic fluctuation rates to obtain M first fluctuation adjustment coefficients;

[0139] Using the flow weight and the load weight, weighted calculation is performed on the M first flow adjustment coefficients and the M first fluctuation adjustment coefficients to obtain M step adjustment coefficients;

[0140] M step size adjustment coefficients are respectively used to adjust and calculate the preset optimization step size to obtain M optimization step sizes.

[0141] In the embodiment of the present application, in order to achieve efficient application of the particle swarm optimization algorithm in the site selection of electric vehicle charging stations, it is necessary to quantify the optimization step size based on the traffic flow scale and fluctuation characteristics of each node.

[0142] Specifically, after obtaining the first traffic rates and first traffic fluctuation rates of M first nodes, the ratio of each first traffic rate to the average of the M first traffic rates is first calculated, and this is used as the M first traffic adjustment coefficients (i.e., first traffic adjustment coefficient = first traffic rate / average of M first traffic rates).

[0143] For example, if M=3, the first traffic rates of the three first nodes A, B, and C are 0.3, 0.2, and 0.5 respectively, and their average is (0.3+0.2+0.5) / 3≈0.333. Then the first flow adjustment coefficient of node A is 0.3 / 0.333≈0.9, the first flow adjustment coefficient of node B is 0.2 / 0.333≈0.6, and the first flow adjustment coefficient of node C is 0.5 / 0.333≈1.5.

[0144] The first flow adjustment coefficient reflects the degree to which each node's traffic flow scale deviates from the overall average. A coefficient greater than 1 indicates that the node's traffic flow is above average, while a coefficient greater than 1 indicates that the node's traffic flow is below average. For example, the traffic flow in residential area C is 50% higher than the average (|1 - 1.5| × 100%).

[0145] Furthermore, the ratio of each first traffic volatility to the average of the M first traffic volatility rates is calculated to obtain M first volatility adjustment coefficients (ie, first volatility adjustment coefficient=first traffic volatility / average of the M first traffic volatility rates).

[0146] For example, if the first traffic fluctuation rates for nodes A, B, and C are 0.8, 0.5, and 0.5, respectively, and their average is (0.8 + 0.5 + 0.5) / 3 = 0.6, then the first fluctuation adjustment coefficient for node A is 0.8 / 0.6 ≈ 1.33, the first fluctuation adjustment coefficient for node B is 0.5 / 0.6 ≈ 0.83, and the first fluctuation adjustment coefficient for node C is 0.5 / 0.6 ≈ 0.83. This coefficient reflects the relative differences in traffic flow stability at each node; a larger coefficient indicates more severe traffic flow fluctuations. For example, the traffic flow fluctuation at shopping mall A is approximately 33% higher than the average (|1 - 1.33| × 100%).

[0147] Furthermore, the obtained flow weight and load weight are used to perform weighted calculation on the M first flow adjustment coefficients and the M first fluctuation adjustment coefficients to obtain M step adjustment coefficients, which are calculated by the formula "step adjustment coefficient = flow weight × first flow adjustment coefficient + load weight × first fluctuation adjustment coefficient".

[0148] For example, if the flow weight is 44.4% and the load weight is 55.6%, taking node A as an example, its step size adjustment coefficient is 44.4% × 0.9 + 55.6% × 1.33 ≈ 1.14. This coefficient comprehensively considers the impact of traffic flow scale and volatility on site selection. The larger the flow weight, the larger the step size adjustment coefficient for high-flow nodes; the larger the load weight, the larger the step size adjustment coefficient for highly volatile nodes.

[0149] Furthermore, the preset optimization step length is adjusted and calculated using M step length adjustment coefficients, respectively, to obtain M optimization step lengths. The specific calculation formula is "optimization step length = step length adjustment coefficient × preset step length", and the value is rounded up.

[0150] For example, if the preset optimization step size is 2 nodes (i.e., each iteration allows a node to move to the second adjacent node), the traffic weight is 44.4%, and the load weight is 55.6%, for Mall A, the step size adjustment factor is 44.4% × 0.9 + 55.6% × 1.33 ≈ 1.14, resulting in an optimization step size of 1.14 × 2 ≈ 2.28, which, after rounding up, is 3 nodes. Therefore, the actual optimization step size for Mall A is 3 nodes.

[0151] Similarly, the step adjustment coefficient of Office Building B is 44.4%×0.6+55.6%×0.83≈0.73, and the optimization step is 0.73×2≈1.46, rounded up to 2 nodes. The actual optimization step of Office Building B is 2 nodes.

[0152] Similarly, the step size adjustment coefficient of residential area C is 44.4%×1.5+55.6%×0.83≈1.11, so the optimization step size is 1.11×2≈2.22, rounded up to 3 nodes, so the actual optimization step size of residential area C is also 3 nodes.

[0153] Furthermore, the M first nodes are updated according to the calculated M optimization step sizes, thereby obtaining M second nodes. That is, each first node moves in the node space according to its corresponding optimization step size.

[0154] For example, the optimization step length of shopping mall A is 3 nodes. During the iteration process, shopping mall A will move from its current position to the adjacent node 3 times in sequence. If it is initially located in the city's commercial district, it may move to the vicinity of a transportation hub; the optimization step length of office building B is 2 nodes, and it will move to the adjacent node 2 times, such as moving from a business concentration area to a surrounding supporting service area; the optimization step length of residential area C is 3 nodes, and it may move to a commercial area near a main road.

[0155] After completing the update from M first nodes to M second nodes, the iterative optimization of the site selection nodes continues. In each iteration, the traffic count set for each node is re-extracted, and the new traffic rate, traffic volatility, and site selection fitness are calculated. The optimization step size is adjusted again based on the new fitness value, and the node position is updated.

[0156] Furthermore, by repeating this process, the node gradually moves to the area with higher site fitness. The iteration stops when the set convergence conditions are met, such as the improvement of site fitness after multiple consecutive iterations is less than the threshold, or when the preset maximum number of iterations is reached.

[0157] Finally, multiple optimal nodes with the greatest site adaptability are obtained and used as the final site optimization results, providing a scientific basis for the layout of electric vehicle charging stations.

[0158] For example, assuming M=3, the initial three first nodes (shopping mall A, office building B, and residential area C) are optimized in the first round to obtain the second nodes A1, B1, and C1; after entering the iterative optimization stage, the node traffic frequency set is re-extracted each time to calculate the new traffic rate, traffic volatility, and site fitness. For example, the new traffic rate of A1 is about 0.33, and the new traffic volatility is about 0.31. Combined with the weights, the fitness is calculated to be 0.53016, and the optimization step size is adjusted accordingly and the node position is updated.

[0159] This process is repeated. When the fitness improvement is less than 0.01 for three consecutive times in the eighth iteration, the iteration is stopped when the convergence condition is met. Finally, the nodes with the largest location fitness, such as A5, B4, and C6, are determined as the location optimization results of electric vehicle charging stations to match regional traffic demand and grid load.

[0160] The embodiments of the present application achieve the following technical effects through the above specific implementation methods:

[0161] This application proposes a method for optimizing the site selection of electric vehicle charging stations. First, based on multiple nodes of the charging station site selection, an OD matrix of the target area is constructed and updated in real time to obtain an OD matrix sequence that reflects the dynamic changes in traffic flow. By analyzing the ratio of the average traffic frequency of the OD matrix sequence to the preset traffic frequency, the flow coefficient is obtained, and a modified traffic weight is generated by combining it with the preset traffic weight. At the same time, the ratio of the discrete parameter of the traffic frequency to the preset discrete parameter is calculated to obtain the flow fluctuation coefficient, and a modified load weight is generated by combining it with the preset load weight. Then, the flow weight and load weight are obtained through weight matching. Further, based on the particle swarm optimization algorithm, the optimization step size is configured according to the flow weight and load weight, and the site selection nodes are iteratively optimized. That is, the node traffic frequency set is extracted to calculate the traffic rate and fluctuation rate, and the site selection fitness is calculated. The optimization step size is dynamically adjusted through weighted calculation of traffic characteristics and weights, and the node position is updated. Finally, the optimal site selection node that takes into account both the scale and fluctuation characteristics of traffic flow is obtained by convergence.

[0162] The method provided in the embodiment of the present application solves the problem of insufficient site selection accuracy caused by insufficient consideration of dynamic changes in traffic flow and grid load impact in traditional charging station site selection. Through the steps of "constructing an OD matrix sequence - configuring traffic and load weights - site selection iteration based on particle swarm optimization", the adaptability of the site selection process to dynamic changes in traffic flow is effectively enhanced, and the scientificity and reliability of the charging station layout plan in complex traffic scenarios are improved.

[0163] Example 2, as shown in the attached Figure 2 As shown, based on the inventive concept of an electric vehicle charging station site selection optimization method provided in Example 1, this application also provides an electric vehicle charging station site selection optimization system, specifically including:

[0164] OD matrix construction and update module 01 is used to construct the OD matrix of the target area based on multiple nodes of charging station site selection, and update it in real time to obtain the OD matrix sequence;

[0165] Traffic load weight configuration module 02, configured to analyze the traffic coefficient and traffic fluctuation coefficient according to the OD matrix sequence, and configure the traffic weight and load weight;

[0166] The weight-driven site selection optimization module 03 is used to perform site selection optimization within multiple nodes based on particle swarm optimization according to the traffic weight and load weight, and obtain multiple optimal nodes as site selection optimization results, wherein the optimization step size is configured according to the traffic weight and load weight for optimization.

[0167] In one embodiment, the OD matrix construction and updating module 01 is further configured to:

[0168] Acquire multiple nodes for charging station site selection;

[0169] Based on traffic monitoring data, collect traffic data between multiple nodes within a preset time range;

[0170] Construct an OD matrix based on the traffic data between multiple nodes;

[0171] OD matrices within multiple preset time ranges are updated and constructed in real time to obtain OD matrix sequences.

[0172] In one embodiment, the traffic load weight configuration module 02 is further configured to:

[0173] According to the OD matrix sequence, the average traffic frequency is calculated;

[0174] Calculating a ratio of the average traffic number to a preset traffic number as a flow coefficient, and multiplying the ratio by a preset flow weight to obtain a modified flow weight;

[0175] Calculating the traffic frequency discrete parameter of the OD matrix sequence;

[0176] Calculating a ratio of the traffic frequency discrete parameter to a preset traffic frequency discrete parameter as a flow fluctuation coefficient, and multiplying the ratio by a preset load weight to obtain a modified load weight;

[0177] The flow weight and the load weight are calculated based on the modified flow weight and the modified load weight.

[0178] In one embodiment, the weight-driven site selection optimization module 03 is further configured to:

[0179] Based on particle swarm optimization, according to the preset number of nodes M, M first nodes are randomly selected as site selection nodes;

[0180] According to the OD matrix sequence, extracting M first node traffic frequency sets of M first nodes, calculating M first traffic rates and M first traffic volatility rates, and calculating first location fitness;

[0181] configuring M optimization step lengths according to the M first traffic rates and the M first traffic fluctuation rates, wherein the optimization step length includes the number of node movements;

[0182] Use M optimization steps to update M first nodes and obtain M second nodes;

[0183] Perform iterative optimization of the site selection nodes, and after convergence, obtain multiple optimal nodes with the largest site selection fitness as the site selection optimization results.

[0184] Example 3, as shown in the attached Figure 3 As shown, based on the inventive concept of an electric vehicle charging station site selection optimization method provided in Example 1, the present application further provides a computer device 300, which comprises:

[0185] Memory 310, for storing computer programs;

[0186] The processor 320 is configured to read and execute the computer program 311 .

[0187] When the computer program is executed by the processor, the method for optimizing the location of electric vehicle charging stations in the first embodiment is implemented.

[0188] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0189] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0190] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A method for optimizing the site selection of electric vehicle charging stations, characterized in that: The method comprises: Based on multiple nodes of charging station site selection, the OD matrix of the target area is constructed and updated in real time to obtain the OD matrix sequence; According to the OD matrix sequence, the flow coefficient and flow fluctuation coefficient are analyzed, and the flow weight and load weight are configured, including: According to the OD matrix sequence, the average traffic frequency is calculated; Calculating a ratio of the average traffic number to a preset traffic number as a flow coefficient, and multiplying the ratio by a preset flow weight to obtain a modified flow weight; Calculating the traffic frequency discrete parameter of the OD matrix sequence; Calculating a ratio of the traffic frequency discrete parameter to a preset traffic frequency discrete parameter as a flow fluctuation coefficient, and multiplying the ratio by a preset load weight to obtain a modified load weight; Calculating the flow weight and the load weight according to the modified flow weight and the modified load weight includes: Calculating the ratio of the modified flow weight to the sum of the modified flow weight and the modified load weight to obtain a flow weight; Calculating the ratio of the corrected load weight to the sum of the corrected flow weight and the corrected load weight to obtain a load weight; Based on particle swarm optimization, location optimization is performed within multiple nodes according to the traffic weight and the load weight, and multiple optimal nodes are obtained as location optimization results, wherein the optimization step size is configured according to the traffic weight and the load weight for optimization, including: Based on particle swarm optimization, according to the preset number of nodes M, M first nodes are randomly selected as site selection nodes; According to the OD matrix sequence, M first node traffic frequency sets of M first nodes are extracted, M first traffic rates and M first traffic volatility rates are calculated, and the first location fitness is calculated, including: Extracting M first-node traffic frequency sets of M first nodes according to the OD matrix sequence, wherein each first-node traffic frequency set includes multiple traffic times of the first node in multiple OD matrices; Calculating the ratio of the total traffic frequency of each first node traffic frequency set to the sum of the traffic frequencies in the OD matrix sequence to obtain M first traffic rates; Calculate the maximum deviation between the minimum and maximum traffic times in the M first-node traffic times set respectively, and obtain M first traffic fluctuation rates; Using the flow weight and the load weight, weighted calculation is performed on the difference between the M first traffic rates and 1 minus the M first traffic fluctuation rates to obtain M first node location fitnesses, and the average is calculated to obtain the first location fitness; According to the M first traffic rates and the M first traffic fluctuation rates, M optimization step sizes are configured, wherein the optimization step size includes the number of node movements, including: Calculating a ratio of each first traffic rate to an average of the M first traffic rates to obtain M first flow adjustment coefficients; Calculating a ratio of each first traffic fluctuation rate to an average of the M first traffic fluctuation rates to obtain M first fluctuation adjustment coefficients; Using the flow weight and the load weight, weighted calculation is performed on the M first flow adjustment coefficients and the M first fluctuation adjustment coefficients to obtain M step adjustment coefficients; Using M step adjustment coefficients respectively, the preset optimization step length is adjusted and calculated to obtain M optimization step lengths; Use M optimization steps to update M first nodes and obtain M second nodes; Perform iterative optimization of the site selection nodes, and after convergence, obtain multiple optimal nodes with the largest site selection fitness as the site selection optimization results.

2. The electric vehicle charging station site selection optimization method according to claim 1, characterized in that: Based on multiple nodes of the charging station, the OD matrix of the target area is constructed and updated in real time to obtain the OD matrix sequence, including: Obtain multiple nodes for charging station site selection; Based on traffic monitoring data, collect traffic data between multiple nodes within a preset time range; Construct an OD matrix based on the traffic data between multiple nodes; OD matrices within multiple preset time ranges are updated and constructed in real time to obtain OD matrix sequences.

3. The electric vehicle charging station site selection optimization method according to claim 2, characterized in that: Based on the traffic data between multiple nodes, an OD matrix is ​​constructed, including: According to the traffic data between multiple nodes, the number of traffic between every two nodes is extracted; Based on the number of traffic flows between multiple groups of nodes, an OD matrix is ​​constructed, where each group of nodes includes two nodes.

4. An electric vehicle charging station site selection optimization system, characterized in that: The system is used to execute the electric vehicle charging station site selection optimization method according to any one of claims 1 to 3, and the system includes: The OD matrix construction and update module is used to construct the OD matrix of the target area based on multiple nodes of the charging station site selection, and update it in real time to obtain the OD matrix sequence; A flow load weight configuration module, configured to analyze the flow coefficient and the flow fluctuation coefficient according to the OD matrix sequence, and configure the flow weight and the load weight; The weight-driven site selection optimization module is used to perform site selection optimization within multiple nodes based on the flow weight and load weight based on particle swarm optimization, and obtain multiple optimal nodes as site selection optimization results, wherein the optimization step size is configured according to the flow weight and load weight for optimization.

5. A computer device, characterized in that: The device includes a processor and a memory: memory for storing computer programs; A processor is used to read and execute the computer program, thereby implementing the electric vehicle charging station site selection optimization method according to any one of claims 1 to 3.

Images