Electric vehicle charging station site selection and capacity determination method and system

Abstract

The application provides a method and system for site selection and capacity determination of an electric vehicle charging station. A charging demand point set is constructed, each charging demand point including spatial coordinates and daily charging demand as a demand weight; the number of clusters K is determined, the spatial coordinates are taken as cluster samples, and the demand weight is taken as sample weight, a demand weighted K-means clustering is performed, and the charging demand points are divided into K service partitions; a candidate station set is constructed, the cluster centers corresponding to each service partition are mapped to the candidate stations, and the recommended station sites corresponding to each service partition are obtained; the number and power configuration of charging facilities in each recommended station are determined, and a site selection and capacity determination scheme is formed; the site selection and capacity determination scheme is evaluated; when the evaluation result meets the condition, the corresponding site selection and capacity determination scheme is taken as the final recommended scheme; otherwise, the number of clusters K, the candidate station mapping result, and / or the charging facility configuration are adjusted.

Description

Technical Field

[0001] This invention belongs to the technical fields of electric vehicle charging infrastructure planning, urban transportation energy service network optimization, power load spatial configuration and intelligent optimization algorithm application, specifically involving a method and system for electric vehicle charging station site selection and capacity determination based on demand-weighted K-means clustering. Background Technology

[0002] With the rapid growth in the number of electric vehicles, the layout of public charging facilities has become an important part of the coordinated planning of urban transportation systems and power distribution networks. Inappropriate charging station locations can lead to excessively long detours for users in some areas, increased queuing times, localized power supply capacity shortages, or low station utilization rates. Therefore, charging station planning needs to consider traffic demand, power supply, and engineering construction conditions simultaneously.

[0003] On the demand side of transportation, charging demand is typically related to road traffic flow, parking attraction intensity, resident travel intensity, commercial and office distribution, the proportion of electric vehicles, and the proportion of public charging stations. If station placement is determined solely based on administrative area size or experience, the spatial non-uniformity of demand can be easily overlooked, leading to a misalignment between charging stations and actual charging demand hotspots.

[0004] On the power supply side, the capacity configuration of charging stations is related to transformer margin, power distribution line access capacity, fast and slow charging ratio, and single-pile power. If the site is determined first and then the capacity is estimated independently, it is easy to cause a disconnect between the service targets of the site and the capacity configuration, resulting in some sites having insufficient capacity and some sites being idle for a long time.

[0005] In existing technologies, common methods for charging station site selection mainly include coverage modeling, traffic flow modeling, integer programming, intelligent optimization algorithms, and cluster analysis. Coverage modeling can guarantee demand coverage within a certain service radius, but it is insufficient in expressing differences in demand intensity. Traffic flow modeling predicts charging demand based on actual road traffic flow data and matches site layout, but this method only relies on road segment traffic flow to calculate charging demand, ignoring vehicle attributes, land and power grid constraints, site competition, user preferences, and dynamic changes in traffic flow. This results in high modeling costs, a tendency to overestimate actual charging passenger flow, and poor feasibility of site selection solutions. Integer programming can establish comprehensive objectives for cost, distance, and capacity, but the computational model is complex, often requiring a large number of refined parameters in the preliminary planning stage. The technical approach of using intelligent optimization algorithms for charging station site selection or capacity determination, such as schemes that incorporate coverage radius, construction cost, and load constraints into the charging facility planning model, and publicly available schemes for optimizing electric vehicle charging facility site selection, provides a foundation for optimized charging facility planning. However, there is still room for improvement in demand-weighted clustering partitioning, candidate site mapping, and the coupling of site selection and capacity determination.

[0006] The K-means clustering-based charging station selection scheme uses demand points within a region as samples, divides these demand points into several service areas through clustering, and uses locations near the cluster centers as suggested charging station locations. This scheme is clear in its approach, requires relatively little computation, is suitable for preliminary planning at the regional level, and has become the most widely used basic site selection method in urban charging infrastructure planning. However, traditional K-means primarily divides samples based on geometric distance, failing to fully reflect the differences in charging demand intensity among different demand points, and also failing to address the issues of cluster centers being unsuitable for construction and the disconnect between capacity allocation and actual needs. Specifically, existing traditional K-means site selection methods have the following shortcomings: 1. Treating each demand point with equal weight tends to make cluster centers closer to the geometric center rather than the charging load center. When certain areas have high traffic volume, a high proportion of electric vehicles, or strong demand for public charging, equal-weight clustering cannot fully reflect its influence on site location.

[0007] 2. K-means cluster centers are typically mathematical coordinate points on a continuous plane. These points may fall in the center of a road, inside a building, in green space, water body, or other areas where construction is not permitted. Directly using these centers as the locations of charging stations would make the planning results difficult to implement.

[0008] 3. Some methods separate site location selection and site capacity configuration into two independent steps, lacking a capacity verification process based on service area demand. A reasonable site selection does not necessarily mean that the capacity meets demand, and sufficient capacity configuration does not necessarily mean that the site location is reasonable.

[0009] 4. Existing preliminary planning results often only provide a few recommended sites, lacking closed-loop evaluation of indicators such as coverage, average service distance, capacity satisfaction, and site utilization, resulting in a lack of unified basis for scheme comparison and parameter adjustment. Summary of the Invention

[0010] To address the aforementioned shortcomings in existing technologies, this invention provides a method and system for site selection and capacity determination of electric vehicle charging stations based on demand-weighted K-means clustering.

[0011] According to a first aspect of the present invention, a method for site selection and capacity determination of electric vehicle charging stations is provided, comprising: Obtain basic data for the planning area; Based on the basic data of the planning area, a set of charging demand points is constructed; wherein, each charging demand point includes spatial coordinates and daily charging demand, and the daily charging demand also serves as a demand weight to characterize the intensity of public charging demand. Determine the number of clusters KUsing the spatial coordinates of the charging demand points as clustering samples and the demand weights as sample weights, demand-weighted K-means clustering is performed to divide the charging demand points into groups. K Each service partition is divided into several service partitions, and the cluster center corresponding to each service partition is obtained. Construct a set of candidate sites that meet the preset construction conditions, map the cluster centers corresponding to each service partition to the candidate sites, and obtain the recommended site address corresponding to each service partition; The daily charging demand of each service zone is summarized to obtain the total daily charging demand of each service zone. Based on the total daily charging demand and the preset capacity constraints of the recommended sites, the number and power configuration of charging facilities in each recommended site are determined to form a site selection and capacity determination scheme. The site selection and capacity allocation schemes are evaluated based on the service distance from charging demand points to the corresponding recommended sites, demand coverage, and capacity satisfaction. When the evaluation results meet a preset planning threshold, the corresponding site selection and capacity allocation scheme is adopted as the final recommended scheme. When the evaluation results do not meet the preset planning threshold, the number of clusters is adjusted. K The candidate site mapping results and / or charging facility configuration will be adjusted until the evaluation results meet the preset planning threshold.

[0012] Preferably, constructing a set of charging demand points based on the basic data of the planned area includes: The road nodes, traffic zones, or parking attractions within the planning area are abstracted as charging demand points, and each charging demand point consists of spatial coordinates and demand weights. Construct a set of charging demand points, represented as: In the formula, A set of charging demand points; For the first One charging demand point; n The total number of charging points; and The first The horizontal and vertical coordinates of each charging demand point in a unified plane coordinate system; For the first The daily charging demand of each charging point is used as the demand weight of that charging point. For the first The daily vehicle volume corresponding to each charging demand point; when the charging demand point is a road node or traffic zone, the daily vehicle volume is the measured daily traffic flow or predicted daily traffic flow of the corresponding road or traffic zone; when the charging demand point is a parking attraction point, the daily vehicle volume is the measured daily number of vehicles visiting the corresponding parking attraction point or the predicted daily number of vehicles visiting the corresponding parking attraction point. The proportion of electric vehicles in the daily vehicle fleet; The proportion of electric vehicles that use public charging methods; The average charging capacity per vehicle using public charging methods; This refers to the average battery capacity of the vehicle. The target state of charge after the vehicle is fully charged; Average state of charge before charging the vehicle.

[0013] Preferably, the daily charging demand of the charging point is determined by any of the following methods: Get the i The measured or predicted daily traffic flow of each charging demand point corresponds to a road or traffic area, and the measured or predicted daily traffic flow is used as the daily traffic flow in formula (3). V i Calculate the first i Daily charging demand at each charging point q i ; Get the i Based on the historical charging orders within the spatial range corresponding to each charging demand point, and according to the charging volume of each day's charging orders within a preset statistical period, the first... i Average daily charging demand at each charging point q i ; Using historical charging orders or traffic flow as basic input, and combining at least one auxiliary feature from parking duration and land use type, the data is input into a pre-established charging demand prediction model, which then outputs the first... i Forecast daily charging demand at each charging point q i .

[0014] Preferably, the step of determining the number of clusters K ,include: Based on the number of charging stations allowed to be built in the planned area, a minimum number of candidate clusters is set. K min Up to the maximum number of candidate clusters K max The range of candidate cluster numbers, where the maximum number of candidate clusters is [not specified]. K max The number of charging stations shall not exceed the maximum permitted number for construction within the planned area. For each integer within the range of the number of candidate clusters K Perform demand-weighted K-means clustering separately to obtain KEach service partition and its corresponding cluster center are assigned, and the corresponding weighted clustering objective function value, weighted average service distance, demand coverage, and maximum service distance are calculated. The number of candidate clusters with a demand coverage rate no lower than the preset minimum demand coverage rate and a maximum service distance no greater than the planned service radius. K The number of feasible clusters was determined. Based on the number of candidate clusters K Compare the weighted clustering objective function value and weighted average service distance before and after increasing the number of clusters, in ascending order; where the number of candidate clusters is denoted as . K The weighted clustering objective function value and weighted average service distance at time are respectively J K and The number of clusters is from K Increase to K When +1, the decrease in the weighted clustering objective function value and the weighted average service distance are respectively and When the number of clusters is K Increase to K After adding 1, if the decrease in the weighted clustering objective function value is no greater than a preset objective function improvement threshold and the decrease in the weighted average service distance is no greater than a preset distance improvement threshold, then... K The number of candidate clusters that meet the condition of slowing down the improvement of the indicator; Among the candidate clusters that satisfy the conditions of demand coverage, maximum service distance, and slowing improvement of indicators, the candidate cluster with the smallest value is selected as the target cluster number. K ; If no candidate cluster size satisfies the condition of slowing improvement in the index, then the minimum number of feasible cluster sizes is selected as the target cluster size. K ; If there are no candidate clusters within the range that simultaneously meet the preset minimum coverage requirement and the planned service radius constraints, then the maximum number of candidate clusters, the planned service radius, and the candidate site set are adjusted, and the target number of clusters is redefined. K ; The weighted average service distance and demand coverage are calculated according to formulas (5) and (6), respectively: In the formula, The number of candidate clusters is K Weighted average service distance over time; n The total number of charging points; K The number of candidate clusters; For the first The charging demand point is relative to the first The allocation variable for the i-th cluster, when the i-th cluster... The charging demand point belongs to the first When performing clustering, Select 1 if the value is 1, otherwise select 0, and each charging demand point belongs to only one cluster; For the first Daily charging demand at each charging point. For the first The charging demand point to the first The Euclidean distance between the cluster centers in a unified planar coordinate system; For the first Each charging demand point to K The minimum distance among the cluster centers; For demand coverage; To plan the service radius; This is an indicator function; it takes the value 1 if the condition within the parentheses is true, and 0 otherwise. The maximum service distance is the distance corresponding to all charging demand points. The maximum value in; K min To minimize the number of candidate clusters; K max This represents the maximum number of candidate clusters.

[0015] Preferably, the demand-weighted K-means clustering method divides the charging demand points into... K Each service partition is assigned a cluster center, and the cluster centers for each service partition are obtained, including: Using the spatial coordinates of charging demand points as clustering samples and the daily charging demand of charging demand points as sample demand weights, demand-weighted K-means clustering is performed. The clustering objective is set as minimizing the weighted sum of squared distances. The allocation variables are determined based on the distances from charging demand points to each cluster center. Then, the cluster centers are updated according to the demand weights of the charging demand points. The demand-weighted K-means clustering process can be represented as follows: In the formula, J The objective function for weighted clustering; n The total number of charging points; K This represents the number of clusters. z ij For the first i The charging demand point is relative to the first j The binary assignment variable for the i-th cluster, when the th cluster... i The charging demand point was allocated to the first j When performing clustering, z ij =1, otherwise zij =0, and each charging demand point belongs to only one cluster; q i For the first i The daily charging demand of each charging point is used as the demand weight of the corresponding charging point. For the first i Location vectors of charging demand points; For the first j Cluster centers; For the first k Cluster centers; The preset iteration convergence threshold; For the first t The weighted clustering objective function value after the next iteration; For the first t The weighted clustering objective function value after -1 iterations; Repeat the charging demand point allocation process shown in formula (8) and the cluster center update process shown in formula (9). When the weighted clustering objective function value of two adjacent iterations satisfies the convergence condition shown in formula (10), or when the number of iterations reaches the preset maximum number of iterations, stop the iteration. Through the above iterative process, a value matching the intensity of charging demand is obtained. K Each service partition and its corresponding cluster center.

[0016] Preferably, the step of constructing a set of candidate sites that meet preset construction conditions, mapping the cluster centers corresponding to each service partition to the candidate sites, obtaining recommended sites for each service partition, and summarizing the daily charging demand of charging demand points in each service partition to obtain the total daily charging demand of each service partition, and determining the number and power configuration of charging facilities in each recommended site based on the total daily charging demand and the preset capacity constraints of the recommended sites to form a site selection and capacity determination scheme, includes: Construct a candidate site set S Candidate site set S It mainly consists of plots of land that meet basic construction conditions, parking lots, bus stations, commercial supporting spaces and / or locations near existing power distribution facilities; For candidate site set S The first in m Each candidate site is given a constructive variable. γ m ; when the first m When the land use, parking conditions, road accessibility, safety distance, and power distribution access conditions of each candidate site all meet the corresponding preset construction conditions. γ m =1, otherwise γ m =0; For each cluster center, perform candidate site mapping, while satisfying the constructibility variables. γ m Among the candidate sites with a value of 1, the candidate site with the lowest overall mapping cost is selected as the recommended site for the corresponding service partition. Then: In the formula, For the first j The recommended candidate sites for each service partition are the recommended site addresses; S For the candidate site set; s m For candidate site set S The first in m One candidate site; γ m For the first m Constructability variables for each candidate site; C jm For the first j The cluster centers and the first m The overall mapping cost between candidate sites; D j For the first j Daily charging demand for each service zone; z ij For the first i The charging demand point is relative to the first j Assignment variables for each cluster; q i For the first i Daily charging demand at each charging point. Q j For the first j The daily service capacity of each recommended site; T d Effective service hours per day; A correction factor is used for charging piles; and These are the rated power of fast charging piles and slow charging piles, respectively; and The first j The number of fast charging piles and slow charging piles configured at each recommended site; In determining the first jWhen determining the number of fast charging piles and slow charging piles for each recommended site, the preset fast charging ratio is first determined based on the demand density, land use type, and user charging timeliness requirements of the corresponding service area. Under the constraints of the maximum number of charging piles allowed to be installed at the recommended site and the power distribution access capacity, non-negative integer configuration schemes for the number of fast charging piles and slow charging piles are enumerated, and the daily service capacity corresponding to each configuration scheme is calculated according to formula (13). Configuration schemes that meet the capacity constraints shown in formula (14) are selected, and the configuration scheme with the smallest difference between the actual fast charging ratio and the preset fast charging ratio (not exceeding the preset ratio deviation threshold) and the daily service capacity and the daily charging demand of the service area is selected as the site selection and capacity determination scheme. When multiple configuration schemes meet the above conditions, the configuration scheme with the smallest total number of charging piles is selected as the site selection and capacity determination scheme.

[0017] Preferably, the candidate site mapping adopts the rule of minimizing the comprehensive mapping cost; wherein, the comprehensive mapping cost is determined based on at least one of the following: distance between the cluster center and the candidate site, construction cost of the candidate site, power distribution access cost, road accessibility, and parking resources; or, through a multi-indicator weighting method, the distance, construction cost, power distribution access cost, road accessibility, and parking resources are respectively processed into dimensionless values, and the road accessibility and parking resources are converted into cost-based indicators, and the comprehensive mapping cost is calculated according to the preset weights corresponding to each indicator; among the candidate sites that meet the constructability conditions, the candidate site with the minimum comprehensive mapping cost is selected as the recommended site for the corresponding service partition; The fast charging ratio is calculated as follows: In the formula, For the first Demand density of each service partition; For the first Daily charging demand for each service zone; For the first The area of ​​each service zone; For the first The fast charging configuration ratio of each station is used as a fast charging ratio indicator. N h,j and N l,j The first j The recommended number of fast charging stations and slow charging stations configured at each site.

[0018] Preferably, the evaluation of the site selection and capacity allocation scheme based on the service distance from the charging demand point to the corresponding recommended site, the demand coverage, and the capacity satisfaction includes: Calculate the capacity fulfillment ratio, weighted average service distance from charging demand point to the corresponding recommended site, maximum service distance, demand coverage rate, and site capacity utilization rate for each recommended site. When the daily service capacity of each recommended site is not lower than the daily charging demand of the corresponding service area, the capacity fulfillment ratio is not lower than the preset capacity fulfillment ratio threshold, the weighted average service distance is not greater than the preset average service distance threshold, the maximum service distance is not greater than the planned service radius, the demand coverage rate is not lower than the preset minimum demand coverage rate, and the site capacity utilization rate is within the preset utilization rate range, the site selection and capacity setting scheme is determined to meet the planning requirements. If the daily service capacity of any recommended site is lower than the daily charging demand of the corresponding service area, or if the capacity fulfillment ratio, weighted average service distance, maximum service distance, demand coverage, or site capacity utilization does not meet the corresponding preset planning threshold, then the number of clusters will be adjusted. K The candidate site mapping rules and / or fast / slow charging configuration parameters are then used to re-execute clustering, site mapping, capacity configuration, and scheme evaluation; wherein, the fast / slow charging configuration parameters include the rated power of the fast charging pile. P h Rated power of slow charging pile P l , No. j Number of fast charging stations at each recommended site N h,j and the number of slow charging stations N l,j .

[0019] Preferably, the evaluation indicators for the scheme also include: construction cost, operation and maintenance cost, carbon emissions, user waiting time, land use efficiency, distribution network safety margin and service fairness; After selecting feasible site selection and capacity determination schemes that meet the preset planning thresholds based on the capacity fulfillment ratio, weighted average service distance, maximum service distance, demand coverage, and site capacity utilization, a comprehensive evaluation of each feasible site selection and capacity determination scheme is conducted using the construction cost, operation and maintenance cost, carbon emissions, user waiting time, land use efficiency, distribution network safety margin, and service fairness; wherein: Construction cost, operation and maintenance cost, carbon emissions, and user waiting time are used as cost-type indicators, while land use efficiency, distribution network safety margin, and service fairness are used as benefit-type indicators. Each evaluation indicator is dimensionless to convert it to the same evaluation scale. Non-negative weights are assigned to each evaluation indicator according to urban planning goals, and the sum of the weights of each evaluation indicator is 1. The comprehensive evaluation score of each feasible site selection and capacity determination scheme is calculated based on the dimensionless evaluation indicator values ​​and corresponding weights, and the feasible site selection and capacity determination scheme with the highest comprehensive evaluation score is determined as the final recommended scheme.

[0020] According to another aspect of the present invention, a site selection and capacity determination system for electric vehicle charging stations is provided, comprising: The basic data acquisition module for the planning area is used to acquire basic data for the planning area. A charging demand point set construction module is used to construct a charging demand point set based on the basic data of the planning area. Each charging demand point includes spatial coordinates and daily charging demand, and the daily charging demand also serves as a demand weight to characterize the intensity of public charging demand. Cluster Quantity Determination and Demand-Weighted Clustering Module: This module is used to determine the number of clusters. K Using the spatial coordinates of the charging demand points as clustering samples and the demand weights as sample weights, demand-weighted K-means clustering is performed to divide the charging demand points into groups. K Each service partition is divided into several service partitions, and the cluster center corresponding to each service partition is obtained. The candidate site mapping module is used to construct a set of candidate sites that meet the preset construction conditions, map the cluster centers corresponding to each service partition to the candidate sites, and obtain the recommended site address corresponding to each service partition. The capacity configuration module is used to summarize the daily charging demand of charging demand points in each service zone, obtain the total daily charging demand of each service zone, and determine the number and power configuration of charging facilities in each recommended site based on the total daily charging demand and the preset capacity constraints of the recommended sites, so as to form a site selection and capacity setting scheme. The scheme evaluation and correction module evaluates the site selection and capacity allocation scheme based on the service distance from the charging demand point to the corresponding recommended site, demand coverage, and capacity satisfaction. When the evaluation result meets the preset planning threshold, the corresponding site selection and capacity allocation scheme is adopted as the final recommended scheme; when the evaluation result does not meet the preset planning threshold, the number of clusters is adjusted. K The candidate site mapping results and / or charging facility configuration will be adjusted until the evaluation results meet the preset planning threshold.

[0021] By adopting the above technical solution, the present invention has at least one of the following beneficial effects compared with the prior art: This invention uses demand weights in clustering calculations to transform the cluster center from the geometric center to the demand center, which better reflects the site layout needs of areas with high traffic flow and high public charging ratio, thereby reducing the average service distance and unnecessary detours for users.

[0022] This invention transforms mathematical cluster centers on a continuous plane into candidate sites that meet actual construction conditions through a candidate site mapping step, avoiding the problem of ordinary clustering methods outputting unconstructable coordinate points and improving the engineering feasibility of the planning results.

[0023] This invention directly links service partitions with capacity configuration, so that each recommended site corresponds to a clear service target and demand scale, which helps to reduce the phenomena of insufficient capacity, excess capacity and uneven site utilization.

[0024] This invention uses indicators such as coverage, average weighted distance, capacity satisfaction, demand density, and fast charging ratio to form an evaluation closed loop, enabling planners to adjust the K value, site mapping, and capacity configuration according to changes in the indicators, thereby improving the efficiency of scheme comparison and selection.

[0025] The data requirements of this invention are relatively clear, and it can be used for preliminary planning of urban areas, as well as in scenarios such as the expansion of existing charging station networks, the construction of supporting facilities for bus stations, and the layout of charging facilities in commercial parking lots. Attached Figure Description

[0026] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the spatial distribution of demand points and candidate charging stations in a preferred embodiment of the present invention.

[0027] Figure 2 This is a preferred embodiment of the invention showing the changes in clustering indices under different K values.

[0028] Figure 3 This is the result of charging demand clustering and site mapping when K=5 in a preferred embodiment of the present invention.

[0029] Figure 4 This is a comparison of evaluation indicators for different site selection schemes in a preferred embodiment of the present invention.

[0030] Figure 5 This is a flowchart illustrating the process of a method for site selection and capacity determination of electric vehicle charging stations according to an embodiment of the present invention. Detailed Implementation

[0031] The embodiments of the present invention are described in detail below: These embodiments are implemented based on the technical solution of the present invention, and provide detailed implementation methods and specific operation processes. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention.

[0032] Existing charging station selection schemes based on K-means clustering typically face the following technical challenges: First, how to generate charging demand weights that can be used for clustering calculations from traffic flow, electric vehicle ratio, and average charging capacity; second, how to obtain service zones that more closely reflect the actual load center under conditions of uneven demand spatial distribution; third, how to convert the mathematical cluster centers obtained by the K-means algorithm into actual candidate sites that meet road accessibility, parking conditions, land use conditions, and power distribution access conditions; and fourth, how to couple the service zone demand with the site capacity configuration to form an evaluable, comparable, and feasible charging station planning scheme.

[0033] To address the aforementioned issues, one embodiment of the present invention provides a method for site selection and capacity determination of electric vehicle charging stations based on demand-weighted K-means clustering. This method, based on demand-weighted clustering partitioning, transforms the clustering algorithm from a simple data analysis tool into a charging station planning tool for engineering applications through demand weight construction, demand-weighted clustering, candidate site mapping, capacity configuration, and evaluation correction. Specifically for the preliminary planning of public charging stations in urban areas, it solves the problem of the difficulty in coordinating multi-source data such as road nodes, traffic flow, charging demand intensity, candidate construction sites, and power distribution access conditions for site selection. This method uses charging demand points as the basic object, unifying demand point coordinates, demand intensity, service radius, candidate site constructibility, and site capacity constraints into a single site selection and capacity determination process. Unlike simply providing the charging station equipment structure, this method proposes a charging station planning technique comprised of data processing steps, clustering partitioning steps, candidate site mapping steps, capacity configuration steps, and evaluation correction steps. The result is no longer an abstract coordinate point, but a complete planning result including site location, service partition, partition demand, capacity configuration, and evaluation indicators.

[0034] Specifically, such as Figure 5 As shown, the electric vehicle charging station site selection and capacity determination method provided in this embodiment may include: S1, Obtain basic data for the planning area; S2, based on the basic data of the planning area, constructs a set of charging demand points; each charging demand point includes spatial coordinates and daily charging demand, and the daily charging demand also serves as a demand weight to characterize the intensity of public charging demand. S3, Determine the number of clusters K Using the spatial coordinates of charging demand points as clustering samples and the demand weights as sample weights, demand-weighted K-means clustering is performed to divide the charging demand points into... K Each service partition is divided into several service partitions, and the cluster center corresponding to each service partition is obtained. S4, construct a set of candidate sites that meet the preset construction conditions, map the cluster centers corresponding to each service partition to the candidate sites, and obtain the recommended site address corresponding to each service partition; S5, summarize the daily charging demand of charging demand points in each service zone, obtain the total daily charging demand of each service zone, and determine the number and power configuration of charging facilities in each recommended site based on the total daily charging demand and the preset capacity constraints of the recommended sites, thus forming a site selection and capacity determination scheme. S6. Evaluate the site selection and capacity allocation scheme based on the service distance from the charging demand point to the corresponding recommended site, the demand coverage, and the capacity satisfaction. When the evaluation result meets the preset planning threshold, the corresponding site selection and capacity allocation scheme is taken as the final recommended scheme. When the evaluation result does not meet the preset planning threshold, the number of clusters is adjusted. K The candidate site mapping results and / or charging facility configuration are adjusted, and the corresponding steps and evaluations are re-executed until the evaluation results meet the preset planning thresholds.

[0035] In some preferred embodiments, the above-mentioned S1, obtaining basic data of the planning area, may further include: Collect basic data for the planning area. The basic data for the planning area adopts data points that can form spatial coordinates and demand weights, including: planning area boundary, road node or traffic zone location, daily traffic flow, parking attraction points, electric vehicle ratio, public charging ratio, average charging capacity per vehicle, coordinates of candidate construction sites, land feasibility, parking conditions, road accessibility, power distribution access conditions, parking orders, charging orders, mobile phone signaling hot zones, taxi trajectories, public transport passenger flow and / or road checkpoint traffic.

[0036] In some preferred embodiments, the above-mentioned S2, which constructs a set of charging demand points based on basic data of the planning area, may further include: S21, road nodes, traffic zones or parking attractions within the planning area are abstracted as charging demand points. Each charging demand point consists of spatial coordinates and demand weights. The demand weights are used to represent the intensity of public charging demand for the corresponding charging demand point within the planning period. S22, construct a set of charging demand points, represented as: In the formula, A set of charging demand points; For the first One charging demand point; n The total number of charging points; and The first The horizontal and vertical coordinates of each charging demand point are in a unified plane coordinate system, and both use the same length unit, which is km or m. For the first The daily charging demand of each charging point is used as the demand weight of that charging point, in kWh / d. For the first The daily traffic flow of the road or traffic zone corresponding to each charging demand point is expressed in veh / d. When the charging demand point is a road node or traffic zone, the daily vehicle volume is the measured or predicted daily traffic flow of the corresponding road or traffic zone. When the charging demand point is a parking attraction point, the daily vehicle volume is the measured or predicted daily number of vehicles visiting the corresponding parking attraction point. The proportion of electric vehicles in the daily vehicle fleet; The proportion of electric vehicles that use public charging methods; The average charging capacity per vehicle using public charging methods is expressed in kWh / veh. The average battery capacity of the vehicle is expressed in kWh. The target state of charge after the vehicle is fully charged; Average state of charge before charging the vehicle; , , and All of these are dimensionless parameters.

[0037] In some preferred embodiments, in step S22 above, the daily charging demand of the charging point can also be determined by any of the following methods: The first method is to obtain the... i The measured or predicted daily traffic flow of each charging demand point corresponds to the road or traffic area, and the measured or predicted daily traffic flow is used as the daily traffic flow in formula (3). Calculate the first i Daily charging demand at each charging point ; The second method is to obtain the... i Based on the historical charging orders within the spatial range corresponding to each charging demand point, and according to the charging volume of each day's charging orders within a preset statistical period, the first... i Average daily charging demand at each charging point ; The third approach uses historical charging orders or traffic flow as the basic input, combined with at least one auxiliary feature from parking duration and land use type, and inputs it into a pre-established charging demand prediction model. The charging demand prediction model then outputs the first... i Forecast daily charging demand at each charging point Land use type is used to characterize the main land use functions within the spatial range corresponding to charging demand points, including at least one of residential land, commercial service land, office land, industrial land, public service facility land, transportation facility land, and mixed land; land use type is obtained through at least one of urban territorial spatial planning data, land use status data, geographic information system land use classification data, point of interest data, or remote sensing image interpretation results.

[0038] Daily charging demand As the first i The demand weight of each charging demand point.

[0039] In some preferred embodiments, S3 above determines the number of clusters. K It can also further include: S311, based on the number of charging stations allowed to be built in the planned area, sets the minimum number of candidate clusters. K min Up to the maximum number of candidate clusters K max The range of candidate cluster numbers, where the maximum number of candidate clusters is [not specified]. K max The number of charging stations shall not exceed the maximum permitted number for construction within the planned area. S312, for each integer within the range of candidate cluster numbers K Perform demand-weighted K-means clustering separately to obtain K Each service partition and its corresponding cluster center are assigned, and the corresponding weighted clustering objective function value, weighted average service distance, demand coverage, and maximum service distance are calculated. S313, select the number of candidate clusters whose demand coverage is not lower than the preset minimum demand coverage and whose maximum service distance is not greater than the planned service radius. K The number of feasible clusters was determined. S314, based on the number of candidate clusters K Compare the weighted clustering objective function value and weighted average service distance before and after increasing the number of clusters, in ascending order; Where the number of candidate clusters is ? K The weighted clustering objective function value and weighted average service distance at time are respectively J K and The number of clusters is from K Increase to K When +1, the decreases in the weighted clustering objective function value and the weighted average service distance are respectively and S315, when the number of clusters is... K Increase toK After adding 1, if the decrease in the weighted clustering objective function value is no greater than the preset objective function improvement threshold, and the decrease in the weighted average service distance is no greater than the preset distance improvement threshold, then... K The number of candidate clusters that meet the condition of slowing down the improvement of the indicator; S316. Among the candidate clusters that meet the conditions of demand coverage, maximum service distance, and slowing improvement of indicators, select the candidate cluster with the smallest value as the target cluster number. K ; S317 If no candidate cluster size satisfies the condition of slowing improvement in the index, then the minimum number of feasible cluster sizes is selected as the target cluster size. K ; S318. If there are no candidate clusters within the range that simultaneously meet the preset minimum coverage requirement and the planned service radius constraints, then adjust the maximum number of candidate clusters, the planned service radius, and the candidate site set, and redetermine the target number of clusters. K .

[0040] Furthermore, the weighted average service distance and demand coverage are calculated according to formulas (5) and (6), respectively: In the formula, The number of candidate clusters is K Weighted average service distance over time; n The total number of charging points; K The number of candidate clusters; For the first The charging demand point is relative to the first The allocation variable for the i-th cluster, when the i-th cluster... The charging demand point belongs to the first When performing clustering, Select 1 if the value is 1, otherwise select 0, and each charging demand point belongs to only one cluster; For the first Daily charging demand at each charging point. For the first The charging demand point to the first The Euclidean distance between the cluster centers in a unified planar coordinate system; For the first Each charging demand point to K The minimum distance among the cluster centers; For demand coverage; To plan the service radius; This is an indicator function; it takes the value 1 if the condition within the parentheses is true, and 0 otherwise. The maximum service distance is the distance corresponding to all charging demand points. The maximum value in;K min To minimize the number of candidate clusters; K max This represents the maximum number of candidate clusters.

[0041] In some preferred embodiments, S3 above performs demand-weighted K-means clustering to divide charging demand points into... K The system can further include: (1) service partitions, and obtaining the cluster centers corresponding to each service partition; S321, using the spatial coordinates of charging demand points as clustering samples and the daily charging demand of charging demand points as sample weights, perform demand-weighted K-means clustering. S322, the clustering objective is set as minimizing the weighted sum of squared distances, so that charging demand points with higher demand intensity have a greater influence on the location of cluster centers. The corresponding allocation variables are determined based on the distance from each charging demand point to its cluster center, and the cluster centers are updated according to the demand weights of each charging demand point. Therefore, the demand-weighted K-means clustering process is expressed as: In the formula, J The objective function for weighted clustering; n The total number of charging points; K This represents the number of clusters. z ij For the first i The charging demand point is relative to the first j The binary assignment variable for the i-th cluster, when the th cluster... i The charging demand point was allocated to the first j When performing clustering, z ij =1, otherwise z ij =0, and each charging demand point belongs to only one cluster; q i For the first i The daily charging demand of each charging point is used as the demand weight of the corresponding charging point. For the first i Location vectors of charging demand points; For the first j Cluster centers; For the first k Cluster centers; The preset iteration convergence threshold; For the first t The weighted clustering objective function value after the next iteration; For the first t The weighted clustering objective function value after -1 iterations.

[0042] S323, repeat the charging demand point allocation process shown in formula (8) and the cluster center update process shown in formula (9). When the weighted clustering objective function value of two adjacent iterations satisfies the convergence condition shown in formula (10), or the number of iterations reaches the preset maximum number of iterations, stop the iteration.

[0043] S324, through the above iterative process, obtains a value that matches the charging demand intensity. K Each service partition and its corresponding cluster center. In some preferred embodiments, in S4 above, a set of candidate sites that meet preset construction conditions is constructed, and the cluster centers corresponding to each service partition are mapped to the candidate sites to obtain recommended sites for each service partition. In S5 above, the daily charging demand of charging demand points within each service partition is summarized to obtain the total daily charging demand of each service partition. Based on the total daily charging demand and the preset capacity constraints of the recommended sites, the number and power configuration of charging facilities in each recommended site are determined to form a site selection and capacity determination scheme, including: S41, Construct a candidate site set S Candidate site set S It mainly consists of plots of land that meet basic construction conditions, parking lots, bus stations, commercial supporting spaces and / or locations near existing power distribution facilities; S42, for the candidate site set S The first in m Each candidate site is given a constructive variable. γ m When the first m When the land use, parking conditions, road accessibility, safety distance, and power distribution access conditions of each candidate site all meet the corresponding preset construction conditions. γ m =1, otherwise γ m =0; S51, perform candidate site mapping for each cluster center, while satisfying the constructibility variables. γ m Among the candidate sites with a value of 1, the candidate site with the lowest overall mapping cost is selected as the recommended site for the corresponding service partition. Then: In the formula, For the first j Recommended candidate sites for each service partition; S For the candidate site set; s m For candidate site set S The first in m One candidate site; γ m For the firstm Constructability variables for each candidate site; C jm For the first j The cluster centers and the first m The comprehensive mapping cost between candidate sites is determined based on at least one of the following: distance from cluster center to candidate site, construction cost, power distribution access cost, road accessibility, and parking resources. D j For the first j Daily charging demand for each service zone; z ij For the first i The charging demand point is relative to the first j Assignment variables for each cluster; q i For the first i Daily charging demand at each charging point. Q j For the first j The daily service capacity of each recommended site is expressed in kWh / d. T d The effective service time per day, expressed in hours (h / d). A correction factor is used for charging piles; and These are the rated power of fast charging piles and slow charging piles, respectively, in kW; and The first j The recommended site is configured with the number of fast charging stations and the number of slow charging stations, and both are non-negative integers.

[0044] S52, determine the first jWhen determining the number of fast charging piles and slow charging piles for each recommended site, the preset fast charging ratio is first determined based on the demand density, land use type, and user charging timeliness requirements of the corresponding service area. User charging timeliness requirements characterize user requirements for charging waiting time, charging completion time, and vehicle dwell time, including at least one of the following: maximum acceptable queuing waiting time, expected charging completion time, average parking dwell time, and charging demand per unit time. User charging timeliness requirements are obtained through historical charging order time data, parking order entry and exit time data, vehicle arrival and departure time data, operational vehicle dispatch data, user survey data, or preset public charging service standards. At least one of the following is obtained; under the constraints of the maximum number of charging piles allowed to be installed at the recommended site and the power distribution access capacity, enumerate non-negative integer configuration schemes for the number of fast charging piles and the number of slow charging piles, and calculate the daily service capacity corresponding to each configuration scheme according to formula (13); select the configuration scheme that meets the capacity constraints shown in formula (14), and select the configuration scheme with the smallest difference between the actual fast charging ratio and the preset fast charging ratio and the preset ratio deviation threshold, and the daily service capacity and the daily charging demand of the zone as the site selection and capacity determination scheme; when there are multiple configuration schemes that meet the above conditions, select the configuration scheme with the smallest total number of charging piles as the site selection and capacity determination scheme.

[0045] In some preferred embodiments, in S51 above, the site mapping adopts the rule of minimizing the comprehensive mapping cost. The comprehensive mapping cost is determined based on at least one of the following: the distance between the cluster center and the candidate site, the construction cost of the candidate site, the power distribution access cost, road accessibility, and parking resources; or, through a multi-index weighting method, the distance, construction cost, power distribution access cost, road accessibility, and parking resources are processed into dimensionless values, and the road accessibility and parking resources are converted into cost-based indicators, and the comprehensive mapping cost is calculated according to the preset weights corresponding to each indicator; among the candidate sites that meet the constructability conditions, the candidate site with the minimum comprehensive mapping cost is selected as the recommended site for the corresponding service partition; wherein: The distance between the cluster center and the candidate site is the Euclidean distance between them in a unified plane coordinate system; if the road topology and road traffic data are available, the distance can also be the shortest road network distance between the corresponding location of the cluster center and the entrance / exit of the candidate site. Construction costs are determined based on at least one of the following: land area of ​​candidate sites, land acquisition costs or site rental costs, site renovation costs, civil construction costs, and charging equipment installation costs. Each cost is calculated based on the corresponding quantity of work and the pre-set unit cost. The cost of power distribution access is determined based on at least one of the following: the line length between the candidate site and the accessible power distribution node, the required new power distribution capacity, the unit cost of cables, the cost of transformer configuration, and the cost of power distribution construction. Road accessibility is determined based on at least one of the following: the road class of the road adjacent to the candidate site, the distance between the candidate site and the main road, the number of connecting roads around the candidate site, and the average travel time to the candidate site; Parking resources are determined based on at least one of the following: the number of available parking spaces at the candidate site, the parking lot area, the parking space utilization rate, and the number of parking spaces that can be converted into charging spaces.

[0046] When using a multi-index weighted approach to determine the comprehensive mapping cost, distance, construction cost, power distribution access cost, road accessibility, and parking resources are processed to be dimensionless. Road accessibility and parking resources are converted into cost-type indicators, and the comprehensive mapping cost is calculated based on the preset weights corresponding to each indicator. Each preset weight is a non-negative number and the sum of the weights is 1. Among the candidate sites that meet the constructability conditions, the candidate site with the smallest comprehensive mapping cost is selected as the recommended site for the corresponding service partition.

[0047] In some preferred embodiments, the demand density and fast charging ratio in step S51 above are calculated in the following ways: In the formula, For the first Demand density of each service partition; For the first Daily charging demand for each service zone; For the first The area of ​​each service zone; For the first The fast charging configuration ratio of each station is used as a fast charging ratio indicator. N h,j and N l,j The first j The recommended number of fast charging stations and slow charging stations configured at each site.

[0048] Among them, demand density is used to identify service zones with different charging demand intensities, and fast charging ratio is used to guide the configuration of fast and slow charging capacity structure for each recommended site.

[0049] In some preferred embodiments, the above-mentioned S6, which evaluates the site selection and capacity allocation scheme based on the service distance from the charging demand point to the corresponding recommended site, the demand coverage, and the capacity satisfaction, may further include: S61, calculate the capacity fulfillment ratio, weighted average service distance from the charging demand point to the corresponding recommended site, maximum service distance, demand coverage rate, and site capacity utilization rate for each recommended site; where: No. j The capacity satisfaction ratio of the recommended sites is: Qj / D j , No. j The site capacity utilization rate of the recommended sites is D j / Q j ;No. i The service distance from each charging demand point to its corresponding recommended station is the Euclidean distance between the charging demand point and its corresponding recommended station in a unified plane coordinate system, or the shortest road network distance between them; the weighted average service distance is calculated by weighting the service distance of each charging demand point with its daily charging demand; the maximum service distance is the maximum value among the service distances of all charging demand points; the demand coverage rate is the proportion of the sum of the daily charging demand of charging demand points whose service distance is not greater than the planned service radius to the sum of the daily charging demand of all charging demand points; S62. When the daily service capacity of each recommended site is not lower than the daily charging demand of the corresponding service area, the capacity fulfillment ratio is not lower than the preset capacity fulfillment ratio threshold, the weighted average service distance is not greater than the preset average service distance threshold, the maximum service distance is not greater than the planned service radius, the demand coverage rate is not lower than the preset minimum demand coverage rate, and the site capacity utilization rate is within the preset utilization rate range, the recommended site plan is determined to meet the planning requirements. More preferably, the preset threshold for capacity fulfillment ratio is 1, and the preferred range for capacity fulfillment ratio is 1.05 to 1.20; the preferred range for planned service radius is 1.5 to 3.0 km, and more preferably 2.0 km; the preferred range for preset average service distance threshold is 0.8 to 1.5 km; the preferred range for preset minimum demand coverage rate is 95% to 100%, and more preferably 100%; the preferred range for site capacity utilization rate is 0.60 to 0.90; the above planning thresholds are adjusted according to the area of ​​the planned region, road network structure, charging demand density, and the number of charging stations allowed to be built; S63. If the daily service capacity of any recommended site is lower than the daily charging demand of the corresponding service area, or if the capacity fulfillment ratio, weighted average service distance, maximum service distance, demand coverage, or site capacity utilization does not meet the corresponding preset planning threshold, then the number of clusters will be adjusted. K Candidate site mapping rules and / or fast / slow charging configuration parameters are then determined, and clustering, site mapping, capacity configuration, and scheme evaluation are re-executed; among which, the fast / slow charging configuration parameters include the rated power of the fast charging pile. P h Rated power of slow charging pile P l , No. j Number of fast charging stations at each recommended site N h,j and the number of slow charging stations Nl,j .

[0050] Further preferably, low demand density thresholds and high demand density thresholds are determined based on the distribution of demand density across all service partitions; preferably, the 25th percentile of the demand density across all service partitions is used as the low demand density threshold, and the 75th percentile is used as the high demand density threshold; when the 25th percentile is used as the high demand density threshold... j When the demand density of a service partition is not lower than the high demand density threshold, it is identified as a high-intensity power replenishment area; when the demand density is lower than the high demand density threshold but not lower than the low demand density threshold, it is identified as a medium-intensity power replenishment area; when the demand density is lower than the low demand density threshold, it is identified as a low-intensity power replenishment area. More preferably, for high-intensity charging areas, the configuration ratio of fast charging piles will be increased. α j The preferred setting is 0.60–0.80; for areas with moderate energy replenishment, the configuration ratio of fast charging stations should be adjusted accordingly. α j The preferred setting is 0.40 to 0.60; for areas with low-intensity energy replenishment, the configuration ratio of fast charging piles should be adjusted. α j The preferred setting is 0.20 to 0.40. Under the conditions of meeting the constraints of the recommended daily service capacity of the site, the maximum number of charging piles, and the power distribution access capacity, the number of fast charging piles and slow charging piles are adjusted so that the actual configuration ratio of fast charging piles is within the corresponding preferred range.

[0051] In some preferred implementation scenarios, the evaluation indicators in S6 above can further include construction costs, operation and maintenance costs, carbon emissions, user waiting time, land use efficiency, distribution network safety margin, and service fairness; among which: S64, after selecting feasible site selection and capacity determination schemes that meet preset planning thresholds based on capacity fulfillment ratio, weighted average service distance, maximum service distance, demand coverage, and site capacity utilization, comprehensively evaluates each feasible site selection and capacity determination scheme using construction cost, operation and maintenance cost, carbon emissions, user waiting time, land use efficiency, distribution network safety margin, and service fairness; among which: Construction costs are determined based on at least one of the following: land acquisition or site rental costs for the recommended sites, site renovation costs, civil construction costs, charging equipment purchase and installation costs, and power distribution access costs. The operation and maintenance cost is determined based on at least one of the following: equipment maintenance costs, labor management costs, site rental costs, and power loss costs of the recommended site within the preset operating cycle; Carbon emissions are determined by multiplying the predicted charging volume of the recommended site within a preset operating cycle with the corresponding electricity carbon emission factor. User waiting time is calculated using a queuing model based on the number of charging vehicles arriving at the recommended station per unit time, the average service rate of a single charging pile, and the number of charging piles configured at the recommended station. Land use efficiency is determined based on the ratio of the daily service capacity of the recommended sites to the area occupied by the recommended sites; The safety margin of the distribution network is determined by the ratio of the difference between the available distribution capacity of the recommended site's access distribution node and the predicted maximum charging load of the recommended site to the available distribution capacity. Service fairness is determined based on the degree of dispersion between the weighted average service distance, demand coverage, or site capacity utilization of different service zones. The smaller the dispersion, the higher the service fairness.

[0052] S65 uses construction cost, operation and maintenance cost, carbon emissions, and user waiting time as cost-type indicators, and land use efficiency, distribution network safety margin, and service fairness as benefit-type indicators. Each evaluation indicator is dimensionless to ensure they are converted to the same evaluation scale. Non-negative weights are assigned to each evaluation indicator based on urban planning objectives, with the sum of the weights being 1. The comprehensive evaluation score for each feasible site selection and capacity determination scheme is calculated based on the dimensionless evaluation indicator values ​​and their corresponding weights, and the feasible site selection and capacity determination scheme with the highest comprehensive evaluation score is selected as the recommended scheme.

[0053] By using the above evaluation indicators, we can comprehensively compare and select the best among multiple feasible site selection and sizing schemes that meet the basic planning thresholds.

[0054] The electric vehicle charging station site selection and capacity determination method provided in the above embodiments of the present invention includes six main steps in its overall process: regional data collection, charging demand point construction, cluster number determination, demand-weighted clustering, candidate site mapping, capacity configuration and scheme evaluation. The data are associated between each step through demand point number, spatial coordinates, partition number and candidate site number.

[0055] Based on the same inventive concept, another embodiment of the present invention provides a site selection and capacity determination system for electric vehicle charging stations based on demand-weighted K-means clustering.

[0056] Specifically, the electric vehicle charging station site selection and capacity determination system provided in this embodiment may include: The planning area basic data collection module is used to acquire basic data of the planning area. This basic data may include: the boundary of the planning area, the location of charging demand points, the daily vehicle scale, the proportion of electric vehicles, the proportion of public charging, candidate site information, and power distribution access conditions. The charging demand point set construction module constructs a set of charging demand points based on the basic data of the planning area. Each charging demand point includes spatial coordinates and daily charging demand, and the daily charging demand also serves as a demand weight to characterize the intensity of public charging demand. Cluster Quantity Determination and Demand-Weighted Clustering Module: This module is used to determine the number of clusters. K Using the spatial coordinates of charging demand points as clustering samples and the demand weights as sample weights, demand-weighted K-means clustering is performed to divide the charging demand points into... K Each service partition is divided into several service partitions, and the cluster center corresponding to each service partition is obtained. The candidate site mapping module is used to construct a set of candidate sites that meet the preset construction conditions, map the cluster centers corresponding to each service partition to the candidate sites, and obtain the recommended site address corresponding to each service partition. The capacity configuration module is used to summarize the daily charging demand of charging demand points in each service zone, obtain the total daily charging demand of each service zone, and determine the number and power configuration of charging facilities in each recommended site based on the total daily charging demand and the preset capacity constraints of the recommended sites, thus forming a site selection and capacity setting scheme. The scheme evaluation and correction module evaluates the site selection and capacity allocation scheme based on the service distance from the charging demand point to the corresponding recommended site, demand coverage, and capacity satisfaction. When the evaluation result meets the preset planning threshold, the corresponding site selection and capacity allocation scheme is adopted as the final recommended scheme; when the evaluation result does not meet the preset planning threshold, the number of clusters is adjusted. K The candidate site mapping results and / or charging facility configuration will be adjusted until the evaluation results meet the preset planning threshold.

[0057] It should be noted that the steps in the method provided by the present invention can be implemented using corresponding modules, devices, units, etc. in the system. Those skilled in the art can refer to the technical solution of the method to realize the composition of the system. That is, the embodiments in the method can be understood as preferred examples for building the system, and will not be elaborated here.

[0058] The following example further illustrates the application effect of the technical solution provided by the above embodiments of the present invention in a specific task.

[0059] This specific application example uses simulation verification data of road nodes, traffic flow, candidate charging sites, and power distribution access conditions in the planning area. During verification, road nodes or traffic zones are first converted into charging demand points, and demand weights are calculated based on traffic flow, the proportion of electric vehicles, the proportion of public charging, and the average charging capacity. A demand-weighted K-means clustering method is then employed.

[0060] like Figure 1The diagram shown is a spatial distribution diagram of demand points and candidate charging stations.

[0061] During the simulation, the number of clusters was changed. It also calculates the weighted objective function, average weighted service distance, and coverage. Typically, While increasing the coverage area reduces the service distance, it increases the number of buildings and the cost of capacity configuration. Therefore, the elbow method and coverage threshold can be used to determine a solution that balances service performance and construction costs.

[0062] After clustering, each cluster center is mapped to a candidate site that meets the construction constraints, and the daily charging demand and service capacity of recommended sites for each service zone are calculated. If the capacity of a site is less than the demand of a zone, it is corrected by increasing the number of charging piles, adjusting the fast and slow charging ratio, or re-dividing the service zones.

[0063] The simulation results can output recommended site locations, site service zones, demand for each zone, site capacity configuration, coverage, average service distance, and capacity satisfaction, proving that the technical solution provided by the above embodiments of the present invention can form complete charging station site selection and capacity determination results, and has engineering application feasibility.

[0064] different Changes in clustering indices under different values ​​are as follows Figure 2 As shown. Figure 2 This indicates that as the number of clusters increases, both the objective function and the destination distance decrease; when When the value exceeds 5, the improvement in the indicator gradually decreases. Considering... When the value is 5, the service coverage reaches 100%, and the maximum destination distance meets the service radius constraint. This example will... =5 is used as the final number of clusters.

[0065] The charging demand clustering and site mapping results at time =5 are as follows: Figure 3 As shown. Figure 3 This intuitively reflects the spatial clustering results of demand points under the condition of the final clustering quantity. The demand points within each service partition are relatively concentrated in space, and the cluster center can represent the demand center of the corresponding partition. After candidate site mapping, the final selected sites can cover the charging demand distribution in different directions of the study area, indicating that the demand-weighted clustering method used in the technical solution provided by the above embodiments of the present invention can identify the location of regional charging demand clusters and form clear candidate site selection results. Figure 3 The mapping relationship between cluster centers and candidate charging stations is given. The cluster centers are not directly used as the final site, but are mapped to candidate sites that meet the construction constraints.

[0066] This specific application example selects three comparison schemes: uniform distribution, traffic flow priority distribution, and random distribution. Each scheme selects 5 charging stations to ensure comparability with the clustering partitioning scheme (K-means clustering partitioning scheme) in this example in terms of the number of stations. Figure 4 As can be seen, the average weighted distance to the station for the clustering and zoning scheme in this example is 0.974km, achieving 100% service coverage. The maximum distance to the station is 1.838km, satisfying the set service radius constraint of 2.0km. The average weighted distance to the station for the uniform distribution scheme is 0.964km, slightly lower than that of the clustering and zoning scheme in this example, but its service coverage is only 97.47%, and the maximum distance to the station reaches 2.088km, exceeding the service radius constraint. This indicates that this scheme has the problem of insufficient coverage of local demand points. Therefore, the uniform distribution scheme cannot be judged to be better solely based on the average distance. The average weighted distance to the station for the traffic flow priority scheme is 0.975km, close to that of the clustering and zoning scheme in this example, but its service coverage is only 91.28%, and the maximum distance to the station reaches 2.220km. This result shows that simply deploying charging stations based on high traffic flow nodes can easily lead to the concentration of stations in local high-traffic areas, resulting in insufficient coverage of edge demand points. The randomized site selection scheme has an average weighted distance to the station of 1.227 km, a service coverage rate of 88.44%, and a maximum distance to the station of 2.184 km. These indicators are generally poor, suggesting that site selection without clear planning basis cannot guarantee stable service performance. In summary, while the clustering and zoning scheme in this example is not the absolute smallest in terms of average weighted distance to the station, it is the only one among the four schemes to simultaneously achieve 100% service coverage, a maximum distance to the station of less than 2.0 km, and 100% capacity fulfillment. For urban public charging station planning, service coverage integrity and the distance to the most unfavorable demand point are of significant engineering importance. If some demand points exceed the service radius, even a low average distance can lead to insufficient charging service in localized areas. Therefore, from the perspectives of coverage integrity, spatial service stability, and engineering feasibility, the clustering and zoning scheme in this example is more suitable as the final site selection scheme.

[0067] The electric vehicle charging station site selection and capacity determination method and system provided in the above embodiments of the present invention are based on demand-weighted K-means clustering. They employ a technique for constructing a set of demand points based on charging demand weights, which transforms traffic flow, electric vehicle ratio, public charging ratio, and average charging capacity per vehicle into clustering weights for each demand point. A demand-weighted clustering partitioning method is used, introducing demand weights into the clustering objective function and center update process to match service partitions with charging demand intensity. A method for mapping cluster centers to constructable candidate sites is employed, converting mathematical cluster centers into actual sites that meet land use, parking, road accessibility, and power distribution access conditions through a set of candidate sites and constructibility variables. A site capacity configuration method based on service partition demand is used, calculating the capacity of fast charging piles, slow charging piles, or other charging facilities for corresponding sites based on the daily charging demand of each service partition. An integrated closed-loop method for site selection, partitioning, capacity determination, and evaluation is used, adjusting the number of clusters, site mapping, and capacity configuration based on coverage, service distance, capacity satisfaction, demand density, and utilization rate.

[0068] Any matters not covered in the above embodiments of the present invention are well-known in the art.

[0069] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various modifications or variations within the scope of the claims, which do not affect the essence of the present invention.

Claims

1. A method for site selection and capacity determination of electric vehicle charging stations, characterized in that, include: Obtain basic data for the planning area; Based on the basic data of the planning area, a set of charging demand points is constructed; wherein, each charging demand point includes spatial coordinates and daily charging demand, and the daily charging demand also serves as a demand weight to characterize the intensity of public charging demand. Determine the number of clusters K Using the spatial coordinates of the charging demand points as clustering samples and the demand weights as sample weights, demand-weighted K-means clustering is performed to divide the charging demand points into... K Each service partition is divided into several service partitions, and the cluster center corresponding to each service partition is obtained. Construct a set of candidate sites that meet the preset construction conditions, map the cluster centers corresponding to each service partition to the candidate sites, and obtain the recommended site address corresponding to each service partition; The daily charging demand of each service zone is summarized to obtain the total daily charging demand of each service zone. Based on the total daily charging demand and the preset capacity constraints of the recommended sites, the number and power configuration of charging facilities in each recommended site are determined to form a site selection and capacity determination scheme. The site selection and capacity allocation schemes are evaluated based on the service distance from charging demand points to the corresponding recommended sites, demand coverage, and capacity satisfaction. When the evaluation results meet a preset planning threshold, the corresponding site selection and capacity allocation scheme is adopted as the final recommended scheme. When the evaluation results do not meet the preset planning threshold, the number of clusters is adjusted. K The candidate site mapping results and / or charging facility configuration will be adjusted until the evaluation results meet the preset planning threshold.

2. The method for site selection and capacity determination of electric vehicle charging stations according to claim 1, characterized in that, The process of constructing a set of charging demand points based on the basic data of the planned area includes: The road nodes, traffic zones, or parking attractions within the planning area are abstracted as charging demand points, and each charging demand point consists of spatial coordinates and demand weights. Construct a set of charging demand points, represented as: In the formula, A set of charging demand points; For the first One charging demand point; n The total number of charging points; and The first The horizontal and vertical coordinates of each charging demand point in a unified plane coordinate system; For the first The daily charging demand of each charging point is used as the demand weight of that charging point. For the first The daily vehicle volume corresponding to each charging demand point; when the charging demand point is a road node or traffic zone, the daily vehicle volume is the measured daily traffic flow or predicted daily traffic flow of the corresponding road or traffic zone; when the charging demand point is a parking attraction point, the daily vehicle volume is the measured daily number of vehicles visiting the corresponding parking attraction point or the predicted daily number of vehicles visiting the corresponding parking attraction point. The proportion of electric vehicles in the daily vehicle fleet; The proportion of electric vehicles that use public charging methods; The average charging capacity per vehicle using public charging methods; This refers to the average battery capacity of the vehicle. The target state of charge after the vehicle is fully charged; Average state of charge before charging the vehicle.

3. The method for site selection and capacity determination of electric vehicle charging stations according to claim 2, characterized in that, The daily charging demand at the charging point is determined by any of the following methods: Get the i The measured or predicted daily traffic flow of each charging demand point corresponds to a road or traffic area, and the measured or predicted daily traffic flow is used as the daily traffic flow in formula (3). V i Calculate the first i Daily charging demand at each charging point q i ; Get the i Based on the historical charging orders within the spatial range corresponding to each charging demand point, and according to the charging volume of each day's charging orders within a preset statistical period, the first... i Average daily charging demand at each charging point q i ; Using historical charging orders or traffic flow as basic input, and combining at least one auxiliary feature from parking duration and land use type, the data is input into a pre-established charging demand prediction model, which then outputs the first... i Forecast daily charging demand at each charging point q i .

4. The method for site selection and capacity determination of electric vehicle charging stations according to claim 1, characterized in that, The determination of the number of clusters K ,include: Based on the number of charging stations allowed to be built in the planned area, a minimum number of candidate clusters is set. K min Up to the maximum number of candidate clusters K max The range of candidate cluster numbers, where the maximum number of candidate clusters is [not specified]. K max The number of charging stations shall not exceed the maximum permitted number for construction within the planned area. For each candidate cluster number within the range of the stated candidate cluster number k Perform demand-weighted K-means clustering separately to obtain k Each service partition and its corresponding cluster center are assigned, and the corresponding weighted clustering objective function value, weighted average service distance, demand coverage, and maximum service distance are calculated. The number of candidate clusters with a demand coverage rate no lower than the preset minimum demand coverage rate and a maximum service distance no greater than the planned service radius. k The number of feasible clusters was determined. Based on the number of candidate clusters k Compare the weighted clustering objective function value and weighted average service distance before and after increasing the number of clusters, in ascending order; where the number of candidate clusters is denoted as . k The weighted clustering objective function value and weighted average service distance at time are respectively J k and The number of clusters is from k Increase to k When +1, the decrease in the weighted clustering objective function value and the weighted average service distance are respectively and When the number of clusters is k Increase to k After adding 1, if the decrease in the weighted clustering objective function value is no greater than a preset objective function improvement threshold and the decrease in the weighted average service distance is no greater than a preset distance improvement threshold, then... k The number of candidate clusters that meet the condition of slowing down the improvement of the indicator; Among the candidate clusters that satisfy the conditions of demand coverage, maximum service distance, and slowing improvement of indicators, the candidate cluster with the smallest value is selected as the target cluster number. K ; If no candidate cluster size satisfies the condition of slowing improvement in the index, then the minimum number of feasible cluster sizes is selected as the target cluster size. K ; If there are no candidate clusters within the range that simultaneously meet the preset minimum coverage requirement and the planned service radius constraints, then the maximum number of candidate clusters, the planned service radius, and the candidate site set are adjusted, and the target number of clusters is redefined. K ; The weighted average service distance and demand coverage are calculated according to formulas (5) and (6), respectively: In the formula, The number of candidate clusters is k Weighted average service distance over time; n The total number of charging points; k The number of candidate clusters; For the first The charging demand point is relative to the first The allocation variable for the i-th cluster, when the i-th cluster... The charging demand point belongs to the first When performing clustering, Select 1 if the value is 1, otherwise select 0, and each charging demand point belongs to only one cluster; For the first Daily charging demand at each charging point. For the first The charging demand point to the first The Euclidean distance between the cluster centers in a unified planar coordinate system; For the first Each charging demand point to k The minimum distance among the cluster centers; For demand coverage; To plan the service radius; This is an indicator function; it takes the value 1 if the condition within the parentheses is true, and 0 otherwise. The maximum service distance is the distance corresponding to all charging demand points. The maximum value in; K min To minimize the number of candidate clusters; K max The maximum number of candidate clusters; K The number of target clusters is determined from the number of candidate clusters.

5. The method for site selection and capacity determination of electric vehicle charging stations according to claim 1, characterized in that, The demand-weighted K-means clustering method divides the charging demand points into... K Each service partition is assigned a cluster center, and the cluster centers for each service partition are obtained, including: Using the spatial coordinates of charging demand points as clustering samples and the daily charging demand of charging demand points as sample demand weights, demand-weighted K-means clustering is performed. The clustering objective is set as minimizing the weighted sum of squared distances. The allocation variables are determined based on the distances from charging demand points to each cluster center. Then, the cluster centers are updated according to the demand weights of the charging demand points. The demand-weighted K-means clustering process can be represented as follows: In the formula, J The objective function for weighted clustering; n The total number of charging points; K This represents the number of clusters. z ij For the first i The charging demand point is relative to the first j The binary assignment variable for the i-th cluster, when the th cluster... i The charging demand point was allocated to the first j When performing clustering, z ij =1, otherwise z ij =0, and each charging demand point belongs to only one cluster; q i For the first i The daily charging demand of each charging point is used as the demand weight of the corresponding charging point. For the first i Location vectors of charging demand points; For the first j Cluster centers; For the first k Cluster centers; The preset iteration convergence threshold; For the first t The weighted clustering objective function value after the next iteration; For the first t The weighted clustering objective function value after -1 iterations; Repeat the charging demand point allocation process shown in formula (8) and the cluster center update process shown in formula (9). When the weighted clustering objective function value of two adjacent iterations satisfies the convergence condition shown in formula (10), or when the number of iterations reaches the preset maximum number of iterations, stop the iteration. Through the above iterative process, a value matching the intensity of charging demand is obtained. K Each service partition and its corresponding cluster center.

6. The method for site selection and capacity determination of electric vehicle charging stations according to claim 1, characterized in that, The process involves constructing a set of candidate sites that meet preset construction conditions, mapping the cluster centers corresponding to each service partition to the candidate sites, obtaining recommended sites for each service partition, and summarizing the daily charging demand of charging demand points within each service partition to obtain the total daily charging demand for each service partition. Based on the total daily charging demand and the preset capacity constraints of the recommended sites, the number and power configuration of charging facilities at each recommended site are determined to form a site selection and capacity allocation scheme. Construct a candidate site set S Candidate site set S It mainly consists of plots of land that meet basic construction conditions, parking lots, bus stations, commercial supporting spaces and / or locations near existing power distribution facilities; For candidate site set S The first in m Each candidate site is given a constructive variable. γ m ; when the first m When the land use, parking conditions, road accessibility, safety distance, and power distribution access conditions of each candidate site all meet the corresponding preset construction conditions. γ m =1, otherwise γ m =0; For each cluster center, perform candidate site mapping, while satisfying the constructibility variables. γ m Among the candidate sites with a value of 1, the candidate site with the lowest overall mapping cost is selected as the recommended site for the corresponding service partition. Then: In the formula, For the first j The recommended candidate sites for each service partition are the recommended site addresses; S For the candidate site set; s m For candidate site set S The first in m One candidate site; γ m For the first m Constructability variables for each candidate site; C jm For the first j The cluster centers and the first m The overall mapping cost between candidate sites; D j For the first j Daily charging demand for each service zone; z ij For the first i The charging demand point is relative to the first j Assignment variables for each cluster; q i For the first i Daily charging demand at each charging point. Q j For the first j The daily service capacity of each recommended site; T d Effective service hours per day; A correction factor is used for charging piles; and These are the rated power of fast charging piles and slow charging piles, respectively; and The first j The number of fast charging piles and slow charging piles configured at each recommended site; In determining the first j When determining the number of fast charging piles and slow charging piles for each recommended site, the preset fast charging ratio is first determined based on the demand density, land use type, and user charging timeliness requirements of the corresponding service area. Under the constraints of the maximum number of charging piles allowed to be installed at the recommended site and the power distribution access capacity, non-negative integer configuration schemes for the number of fast charging piles and slow charging piles are enumerated, and the daily service capacity corresponding to each configuration scheme is calculated according to formula (13). The configuration scheme that meets the capacity constraints shown in formula (14) is selected, and the configuration scheme with the smallest difference between the actual fast charging ratio and the preset fast charging ratio (not exceeding the preset ratio deviation threshold) and the daily service capacity and the daily charging demand of the service area is selected as the site selection and capacity determination scheme. When multiple configuration schemes meet the above conditions, the configuration scheme with the smallest total number of charging piles is selected as the site selection and capacity determination scheme.

7. The method for site selection and capacity determination of electric vehicle charging stations according to claim 6, characterized in that, The candidate site mapping adopts the rule of minimizing the comprehensive mapping cost; wherein, the comprehensive mapping cost is determined based on at least one of the following: distance between the cluster center and the candidate site, construction cost of the candidate site, power distribution access cost, road accessibility, and parking resources; or, through a multi-indicator weighting method, the distance, construction cost, power distribution access cost, road accessibility, and parking resources are respectively processed into dimensionless values, and the road accessibility and parking resources are converted into cost-based indicators, and the comprehensive mapping cost is calculated according to the preset weights corresponding to each indicator; among the candidate sites that meet the constructability conditions, the candidate site with the minimum comprehensive mapping cost is selected as the recommended site for the corresponding service partition; The demand density and fast charging ratio indicators are calculated using the following methods: In the formula, For the first Demand density of each service partition; For the first Daily charging demand for each service zone; For the first The area of ​​each service zone; For the first The fast charging configuration ratio of each station is used as a fast charging ratio indicator. N h,j and N l,j The first j The recommended number of fast charging stations and slow charging stations configured at each site.

8. The method for site selection and capacity determination of electric vehicle charging stations according to claim 1, characterized in that, The evaluation of the site selection and capacity allocation scheme based on the service distance from the charging demand point to the corresponding recommended site, the demand coverage, and the capacity satisfaction includes: Calculate the capacity fulfillment ratio, weighted average service distance from charging demand point to the corresponding recommended site, maximum service distance, demand coverage rate, and site capacity utilization rate for each recommended site. When the daily service capacity of each recommended site is not lower than the daily charging demand of the corresponding service area, the capacity fulfillment ratio is not lower than the preset capacity fulfillment ratio threshold, the weighted average service distance is not greater than the preset average service distance threshold, the maximum service distance is not greater than the planned service radius, the demand coverage rate is not lower than the preset minimum demand coverage rate, and the site capacity utilization rate is within the preset utilization rate range, the site selection and capacity setting scheme is determined to meet the planning requirements. If the daily service capacity of any recommended site is lower than the daily charging demand of the corresponding service area, or if the capacity fulfillment ratio, weighted average service distance, maximum service distance, demand coverage, or site capacity utilization does not meet the corresponding preset planning threshold, then the number of clusters will be adjusted. K The candidate site mapping rules and / or fast / slow charging configuration parameters are then used to re-execute clustering, site mapping, capacity configuration, and scheme evaluation; wherein, the fast / slow charging configuration parameters include the rated power of the fast charging pile. P h Rated power of slow charging pile P l , No. j Number of fast charging stations at each recommended site N h,j and the number of slow charging stations N l,j .

9. The method for site selection and capacity determination of electric vehicle charging stations according to claim 8, characterized in that, The evaluation indicators for the proposed scheme also include: construction cost, operation and maintenance cost, carbon emissions, user waiting time, land use efficiency, distribution network safety margin, and service fairness. After selecting feasible site selection and capacity determination schemes that meet the preset planning thresholds based on the capacity fulfillment ratio, weighted average service distance, maximum service distance, demand coverage, and site capacity utilization, a comprehensive evaluation of each feasible site selection and capacity determination scheme is conducted using the construction cost, operation and maintenance cost, carbon emissions, user waiting time, land use efficiency, distribution network safety margin, and service fairness; wherein: Construction cost, operation and maintenance cost, carbon emissions, and user waiting time are used as cost-type indicators, while land use efficiency, distribution network safety margin, and service fairness are used as benefit-type indicators. Each evaluation indicator is dimensionless to convert it to the same evaluation scale. Non-negative weights are assigned to each evaluation indicator according to urban planning goals, and the sum of the weights of each evaluation indicator is 1. The comprehensive evaluation score of each feasible site selection and capacity determination scheme is calculated based on the dimensionless evaluation indicator values ​​and corresponding weights, and the feasible site selection and capacity determination scheme with the highest comprehensive evaluation score is determined as the final recommended scheme.

10. A site selection and capacity determination system for electric vehicle charging stations, characterized in that, include: The basic data acquisition module for the planning area is used to acquire basic data for the planning area. A charging demand point set construction module is used to construct a charging demand point set based on the basic data of the planning area. Each charging demand point includes spatial coordinates and daily charging demand, and the daily charging demand also serves as a demand weight to characterize the intensity of public charging demand. Cluster Quantity Determination and Demand-Weighted Clustering Module: This module is used to determine the number of clusters. K Using the spatial coordinates of the charging demand points as clustering samples and the demand weights as sample weights, demand-weighted K-means clustering is performed to divide the charging demand points into... K Each service partition is divided into several service partitions, and the cluster center corresponding to each service partition is obtained. The candidate site mapping module is used to construct a set of candidate sites that meet the preset construction conditions, map the cluster centers corresponding to each service partition to the candidate sites, and obtain the recommended site address corresponding to each service partition. The capacity configuration module is used to summarize the daily charging demand of charging demand points in each service zone, obtain the total daily charging demand of each service zone, and determine the number and power configuration of charging facilities in each recommended site based on the total daily charging demand and the preset capacity constraints of the recommended sites, so as to form a site selection and capacity setting scheme. The scheme evaluation and correction module evaluates the site selection and capacity allocation scheme based on the service distance from the charging demand point to the corresponding recommended site, demand coverage, and capacity satisfaction. When the evaluation result meets the preset planning threshold, the corresponding site selection and capacity allocation scheme is adopted as the final recommended scheme; when the evaluation result does not meet the preset planning threshold, the number of clusters is adjusted. K The candidate site mapping results and / or charging facility configuration will be adjusted until the evaluation results meet the preset planning threshold.

Images