A power distribution network constraint-based electric vehicle charging facility planning decision method

Abstract

The application discloses a kind of electric vehicle charging facilities planning decision-making methods based on distribution network constraint, it is related to big data analysis technical field, including following steps: S1, acquisition multi-source planning data and unified processing, generate charging facility planning original data set;S2, extract charging demand, construction resource, existing facility and distribution network constraint feature, generate charging facility planning feature set;S3, screening can be constructed space, generate candidate charging facility site set;S4, construct site demand and distribution network constraint coupling's planning constraint association model;S5, carry out access feasibility check, generate candidate charging facility planning scheme set;S6, input improved MISOCP model and carry out joint optimization;S7, output construction scheme, access scheme, distribution network reconstruction demand and implementation priority.The application adopts planning constraint association model and improved MISOCP model, realizes electric vehicle charging facility and distribution network constraint collaborative planning decision-making.

Description

Technical Field

[0001] This invention relates to the field of big data analytics, and in particular to a planning and decision-making method for electric vehicle charging facilities based on power distribution network constraints. Background Technology

[0002] With the continuous growth of new energy vehicle ownership, the rapid increase in public energy replenishment demand, and the increasing pressure on urban power distribution networks, planning and decision-making technologies for the location, scale, and access schemes of electric vehicle charging facilities have received widespread attention. Existing charging facility planning methods mainly rely on vehicle ownership, historical charging orders, parking resource distribution, or manual experience for site selection and capacity determination, with subsequent verification of access conditions by power distribution network maintenance personnel. However, in practical applications, the following problems commonly exist: Data sources for charging demand, parking resources, the operational status of existing charging facilities, and distribution network operation are scattered, with inconsistent data encoding, spatial location, and temporal references. Existing methods struggle to uniformly link candidate construction sites, charging demand areas, and distribution network access nodes, leading to a disconnect between planning results and actual access conditions. Traditional planning methods often prioritize demand coverage or construction costs, neglecting to consider distribution network constraints such as transformer capacity margin, feeder load rate, voltage deviation, line current carrying capacity, and access distance. This can easily result in redundant construction in high-demand areas or transformer overload, feeder overload, and voltage exceeding limits after access. Existing optimization methods typically solve all candidate schemes directly without fully utilizing access feasibility verification results to narrow the planning feasible region. This leads to invalid schemes participating in optimization calculations, resulting in low planning efficiency. Furthermore, it is difficult to simultaneously output construction schemes, access schemes, distribution network transformation needs, and implementation priorities, affecting the accuracy of collaborative planning decisions for electric vehicle charging facilities and the distribution network.

[0003] Therefore, how to provide a planning and decision-making method for electric vehicle charging facilities based on distribution network constraints is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] One objective of this invention is to propose a planning and decision-making method for electric vehicle charging facilities based on distribution network constraints. This invention combines a planning constraint correlation model that couples site demand with distribution network constraints with an improved MISOCP model. By uniformly modeling charging demand, construction resources, the operational status of existing facilities, and distribution network constraint characteristics, it constructs the correlation between candidate charging facility sites, demand spatial units, and grid access nodes, achieving joint optimization of charging facility construction location, scale, facility type, access nodes, and distribution network modification methods. Furthermore, through access feasibility verification and feasible region shrinkage mechanisms, candidate schemes that do not meet the requirements of access capacity, power supply path, and operational safety boundaries are eliminated in advance, reducing ineffective optimization searches and improving planning efficiency. The final planning result simultaneously meets charging service demand and distribution network safety operation requirements, offering advantages such as high planning accuracy, strong access feasibility, low construction redundancy, and applicability to collaborative planning of urban charging facilities.

[0005] A method for planning and deciding on electric vehicle charging facilities based on distribution network constraints according to an embodiment of the present invention includes the following steps: S1. Collect multi-source planning data related to electric vehicle charging facility planning within the target area, process the multi-source planning data in a unified manner, and generate the original dataset of charging facility planning. S2. Standardize the original dataset of the charging facility planning and extract charging demand features, construction resource features, existing facility features and distribution network constraint features to generate a charging facility planning feature set. S3. Based on the charging facility planning feature set, the available space within the target area is screened to generate a set of candidate charging facility sites; S4. Based on the candidate charging facility site set and the charging facility planning feature set, determine the charging demand distribution information and distribution network access conditions, and construct a planning constraint association model that couples site demand with distribution network constraints, and generate planning constraint association results. S5. Based on the planning constraint association results, verify the access feasibility of the candidate charging facility sites in terms of access capacity, power supply path and operation safety boundary, and generate a set of candidate charging facility planning schemes. S6. Input the set of candidate charging facility planning schemes into the improved MISOCP model, and use the improved MISOCP model to jointly optimize the construction location, construction scale, facility type, access node and distribution network transformation method to generate charging facility planning optimization results; S7. Based on the optimization results of the charging facility planning, determine the construction plan, access plan, power distribution network transformation requirements and implementation priority of the electric vehicle charging facility, and generate electric vehicle charging facility planning decision information.

[0006] Optionally, S1 specifically includes: Acquire charging demand forecast data, vehicle travel data, parking resource data, construction resource data, existing charging facility operation data, power distribution network basic data, power distribution network operation data, and planning and construction data within the target area to generate multi-source planning data; Read the regional affiliation information, site affiliation information, facility affiliation information, equipment affiliation information, and power supply affiliation information from the multi-source planning data, and encode them uniformly according to the preset coding rules to generate a unified identification result; Based on the location description information, spatial coordinate information, power supply range information, and distribution network access location information in the multi-source planning data, the spatial location of the charging demand occurrence area, parking resource location, construction resource location, existing charging facility location, and distribution network access node location is spatially calibrated to generate spatial calibration results. Based on the demand forecast time, vehicle travel time, parking record time, charging facility operation time, power distribution network operation data collection time, and planning and construction time in the multi-source planning data, a unified time benchmark is generated to produce a time stamp result. The unified identification result, the spatial calibration result, and the time stamp result are written into the multi-source planning data, and then linked and organized according to data source, regional affiliation, spatial location, and time stamp to generate the original dataset for charging facility planning.

[0007] Optionally, S2 specifically includes: Read the unified identification results, spatial calibration results, time stamp results, and planning association fields from the original dataset of the charging facility planning, and generate a planning dataset to be processed; The planning dataset to be processed is subjected to field integrity verification, duplicate record merging, abnormal record filtering and format unification processing to generate a valid planning dataset; Based on a unified data format and a unified time granularity, the effective planning dataset is standardized and transformed, and spatiotemporally aligned according to the unified identification results, the spatial calibration results, and the time stamp results to generate a standardized planning dataset. Charging demand features are extracted from the charging demand forecast data and vehicle travel data in the standardized planning dataset, and construction resource features are extracted from the parking resource data and construction resource data in the standardized planning dataset. Existing facility features are extracted based on existing charging facility operation data in the standardized planning dataset, and distribution network constraint features are extracted based on distribution network basic data, distribution network operation data, and planning and construction data in the standardized planning dataset. The charging demand characteristics, construction resource characteristics, existing facility characteristics, and distribution network constraint characteristics are normalized and correlated to generate a charging facility planning feature set.

[0008] Optionally, S3 specifically includes: Read the charging demand characteristics, construction resource characteristics, existing facility characteristics and distribution network constraint characteristics from the charging facility planning feature set, and generate site screening input data; Based on the aforementioned construction resource characteristics, spatial units within the target area that can be used for the construction of electric vehicle charging facilities are determined, and these spatial units are initially screened based on land use status, parking capacity, construction restrictions, and spatial accessibility to generate an initial set of constructable spaces. Based on the charging demand characteristics, determine the charging demand intensity and service coverage corresponding to the initial buildable space set, and remove space units whose charging demand intensity does not meet the preset planning requirements to generate a demand-adaptive space set. Based on the characteristics of the existing facilities, the service overlap status between the demand-adaptive space set and the existing charging facilities is determined, and space units with service overlap exceeding a preset threshold and no need for new construction are eliminated, thereby generating a candidate construction space set. Based on the distribution network constraint characteristics, the candidate construction space set is screened for basic access conditions, and the candidate construction spaces that meet the basic access conditions are determined as candidate charging facility sites, thus generating a candidate charging facility site set.

[0009] Optionally, S4 specifically includes: Read the candidate charging facility site set and extract charging demand features and power distribution network constraint features from the charging facility planning feature set; Based on the charging demand characteristics, determine the charging demand intensity, demand service range and demand time distribution corresponding to different spatial units within the target area, and generate charging demand distribution information. Based on the aforementioned distribution network constraint characteristics, candidate access nodes, power supply paths, capacity margins, voltage margins, and line load margins are determined to generate distribution network access conditions. Using the candidate charging facility sites in the candidate charging facility site set as site nodes, the demand spatial units corresponding to the charging demand distribution information as demand nodes, and the candidate access nodes corresponding to the distribution network access conditions as grid nodes, a planning associated node set is constructed. Based on the spatial coverage and demand matching relationships between the candidate charging facility sites and the demand spatial units, a site demand correlation relationship is established. Based on the matching status of the access distance, power supply path, and access capacity between the candidate charging facility sites and the candidate access nodes, a site access association relationship is established. Based on the capacity margin, voltage margin, and line load margin corresponding to the candidate access nodes, establish the distribution network constraint relationship; The planning-related node set, the site demand relationship, the site access relationship, and the distribution network constraint relationship are coupled and modeled to construct a planning constraint relationship model that couples site demand and distribution network constraints. Based on the planning constraint association model, the demand coverage result, access matching result and distribution network constraint matching result corresponding to the candidate charging facility sites are output, and the planning constraint association result is generated.

[0010] Optionally, the step of coupling and modeling the planning-related node set, the site demand relationship, the site access relationship, and the distribution network constraint relationship to construct a planning constraint relationship model that couples site demand and distribution network constraints specifically involves: The site nodes, demand nodes, and power grid nodes in the planning associated node set are labeled with node types to generate planning node type labeling results; The relationship types of the site demand relationship, the site access relationship, and the distribution network constraint relationship are labeled to generate planning relationship type labeling results; Based on the planning node type labeling results and the planning relationship type labeling results, a three-layer planning association structure including the site layer, demand layer and power grid layer is constructed. Based on the site demand association, the site nodes in the site layer are connected to the demand nodes in the demand layer to generate a site demand coupled substructure; Based on the site access association, the site nodes in the site layer are connected to the power grid nodes in the power grid layer to generate a site access coupling substructure; Based on the distribution network constraint association, the grid nodes in the grid layer are constrained and bound to the corresponding capacity margin, voltage margin and line load margin to generate a grid constraint coupling substructure; The site demand coupling substructure, the site access coupling substructure, and the power grid constraint coupling substructure are indexed in a unified manner to generate a site demand access constraint coupling structure. Based on the site demand access constraint coupling structure, the demand coverage status, access matching status and distribution network constraint status corresponding to each candidate charging facility site are determined, and a planning constraint association model coupling site demand and distribution network constraint is constructed.

[0011] Optionally, S5 specifically includes: Read the demand coverage result, access matching result and distribution network constraint matching result from the planning constraint association result, and generate access feasibility verification input data; Based on the access feasibility verification input data, determine the candidate access nodes, access capacity and access distance corresponding to each candidate charging facility site, and generate access capability verification results; Based on the access matching results, determine the power supply lines, upstream distribution transformers, associated feeders and power supply levels between each candidate charging facility site and the candidate access node, and generate power supply path verification results; Based on the distribution network constraint matching results, determine the capacity margin, voltage margin, line load margin, and power supply reliability status of each candidate access node, and generate the operation safety boundary verification results; Based on the access capacity verification results, the power supply path verification results, and the operation safety boundary verification results, candidate charging facility sites that do not meet the requirements of access capacity, access distance, power supply path connectivity, and operation safety boundary are eliminated to generate a set of feasible candidate sites. Based on the set of feasible candidate sites, determine the accessible nodes, constructable capacity range, optional charging facility types and access constraint status corresponding to each feasible candidate site, and generate a candidate charging facility planning scheme; The candidate charging facility planning schemes corresponding to each feasible candidate site are summarized to generate a set of candidate charging facility planning schemes.

[0012] Optionally, S6 specifically includes: Read the feasible candidate sites, accessible nodes, constructible capacity range, optional charging facility types, and access constraint status from the candidate charging facility planning scheme set, and generate planning optimization input data; Based on the planning optimization input data, determine the construction location variables, construction scale variables, facility type variables, access node variables, and distribution network transformation method variables, and generate an optimization decision variable set; Based on the charging demand characteristics, construction resource characteristics, existing facility characteristics, and distribution network constraint characteristics in the planning optimization input data, a set of planning optimization objectives is generated; Based on the access constraint status, the accessible nodes, and the buildable capacity range, a set of distribution network constraint conditions is generated. Input the set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints into the improved MISOCP model; The improved MISOCP model is used to jointly solve for the construction location, construction scale, facility type, access node and distribution network transformation method to generate a target planning scheme that meets the distribution network constraints. Based on the target planning scheme, determine the construction status, construction capacity, charging facility configuration results, access node selection results, and power distribution network transformation methods for each feasible candidate site, and generate charging facility planning optimization results.

[0013] Optionally, inputting the set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints into the improved MISOCP model specifically involves: The improved MISOCP model includes a planning input module, a variable encoding module, an objective construction module, a second-order cone constraint transformation module, a feasible region shrinkage module, a joint solution module, and a result output module. The planning input module reads the set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints, and extracts construction location variables, construction scale variables, facility type variables, access node variables, distribution network transformation method variables, planning optimization objectives, and distribution network operation constraints. The variable encoding module encodes the construction location variable, facility type variable, access node variable, and distribution network renovation method variable as integer variables, and encodes the construction scale variable as a continuous variable, generating a mixed integer decision variable set. The target construction module constructs a multi-objective planning function set based on the planning optimization target set, which includes demand coverage target, construction cost target, access and transformation cost target, distribution network safety margin target, and facility utilization efficiency target; The second-order cone constraint transformation module performs second-order cone constraint transformation on the voltage safety constraint, line capacity constraint and distribution transformer capacity constraint in the distribution network constraint set, and embeds the access capacity constraint, power supply path constraint and access distance constraint as mixed integer constraints to generate second-order cone constraint conditions. The feasible region shrinking module shrinks the feasible region based on the access constraint status in the candidate charging facility planning scheme set, and shrinks the construction location variables, access node variables and construction scale variables that do not meet the access feasibility verification requirements to generate a shrunken planning feasible region. The joint solution module jointly solves the mixed integer decision variable set, the multi-objective programming function set, the second-order cone constraint condition, and the shrinking planning feasible region to generate an optimized solution result. The result output module outputs the values ​​for construction location, construction scale, facility type, access node, and power distribution network transformation method based on the optimization solution results.

[0014] Optionally, S7 specifically includes: Read the values ​​of construction location, construction scale, facility type, access node, and power distribution network transformation method from the charging facility planning and optimization results, and generate planning decision input data; Based on the values ​​of the construction location and the construction scale, the recommended construction location and recommended construction capacity of electric vehicle charging facilities in the target area are determined, and a construction plan is generated. Based on the value of the facility type, determine the type of charging facility, the number of facilities configured, and the scope of service targets corresponding to each recommended construction location, and generate facility configuration results; Based on the values ​​of the access nodes, the access power point, access line and power supply path corresponding to each recommended construction location are determined, and an access scheme is generated. Based on the value of the power distribution network renovation method, determine the corresponding renovation content for transformer capacity expansion, line renovation, access point expansion and protection configuration adjustment, and generate power distribution network renovation requirements. Based on the construction plan, the access plan, the facility configuration results, and the power distribution network renovation requirements, the recommended construction locations within the target area are sorted in order of implementation to generate an implementation priority. The construction plan, the access plan, the facility configuration results, the power distribution network transformation requirements, and the implementation priority are combined to generate electric vehicle charging facility planning decision information.

[0015] The beneficial effects of this invention are as follows: This invention addresses the problems of scattered multi-source planning data, disconnect between candidate sites and distribution network access conditions, and low efficiency of traditional optimization solutions in electric vehicle charging facility planning by cascading and integrating a planning constraint association model coupling site demand and distribution network constraints with an improved MISOCP model. It employs unified coding, spatial calibration, and time alignment to standardize charging demand forecast data, parking resource data, existing charging facility operation data, and distribution network operation data, achieving a unified expression of charging demand, construction resources, existing facilities, and distribution network constraint characteristics. In the site association stage, a three-layer planning association structure including a site layer, a demand layer, and a grid layer is constructed, coupling and modeling candidate charging facility sites, demand spatial units, and grid access nodes. This is achieved through site demand correlation... The correlation relationships, site access correlation relationships, and distribution network constraint correlation relationships simultaneously characterize the demand coverage status, access matching status, and distribution network constraint status. Addressing the problem of traditional methods emphasizing demand over access and easily causing transformer overload and feeder overload, this invention combines access capacity, power supply path connectivity, capacity margin, voltage margin, and line load margin to eliminate infeasible candidate schemes during the access feasibility verification process. In the planning optimization stage, by improving the MISOCP model and introducing a feasible region shrinkage module and a distribution network operation constraint embedding mechanism, the invention jointly solves for construction location, construction scale, facility type, access nodes, and distribution network transformation methods, reducing invalid search space and improving planning efficiency. Ultimately, it achieves coordinated output of electric vehicle charging facility construction schemes, access schemes, distribution network transformation needs, and implementation priorities. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is an overall flowchart of a planning and decision-making method for electric vehicle charging facilities based on power distribution network constraints proposed in this invention. Figure 2 This is a schematic diagram of the planning constraint association model structure of the electric vehicle charging facility planning decision-making method based on distribution network constraints proposed in this invention; Figure 3 This is a schematic diagram of the improved MISOCP model structure of the electric vehicle charging facility planning and decision-making method based on distribution network constraints proposed in this invention. Detailed Implementation

[0017] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0018] refer to Figure 1-3 A planning and decision-making method for electric vehicle charging facilities based on distribution network constraints includes the following steps: S1. Collect multi-source planning data related to electric vehicle charging facility planning within the target area, process the multi-source planning data in a unified manner, and generate the original dataset of charging facility planning. S2. Standardize the original dataset for charging facility planning and extract charging demand features, construction resource features, existing facility features and distribution network constraint features to generate a charging facility planning feature set. S3. Based on the charging facility planning feature set, screen the buildable space in the target area to generate a set of candidate charging facility sites; S4. Based on the candidate charging facility site set and the charging facility planning feature set, determine the charging demand distribution information and distribution network access conditions, and construct a planning constraint association model that couples site demand with distribution network constraints, and generate planning constraint association results. S5. Based on the planning constraint correlation results, verify the access feasibility of candidate charging facility sites in terms of access capacity, power supply path and operation safety boundary, and generate a set of candidate charging facility planning schemes. S6. Input the set of candidate charging facility planning schemes into the improved MISOCP model. Through the improved MISOCP model, the construction location, construction scale, facility type, access node and distribution network transformation method are jointly optimized to generate charging facility planning optimization results. S7. Based on the optimization results of charging facility planning, determine the construction plan, access plan, power distribution network transformation needs and implementation priorities of electric vehicle charging facilities, and generate planning decision information for electric vehicle charging facilities.

[0019] In this embodiment, S1 specifically refers to: Acquire charging demand forecast data, vehicle travel data, parking resource data, construction resource data, existing charging facility operation data, power distribution network basic data, power distribution network operation data, and planning and construction data within the target area to generate multi-source planning data; Read the regional affiliation information, site affiliation information, facility affiliation information, equipment affiliation information and power supply affiliation information from the multi-source planning data, and encode them uniformly according to the preset coding rules to generate a unified identification result; Based on the location description information, spatial coordinate information, power supply range information and distribution network access location information in the multi-source planning data, the spatial location of the charging demand area, parking resource location, construction resource location, existing charging facility location and distribution network access node location is spatially calibrated, and spatial calibration results are generated. Based on the demand forecast time, vehicle travel time, parking record time, charging facility operation time, power distribution network operation data collection time, and planning and construction time in the multi-source planning data, the time base is unified to generate time stamp results; The unified identification results, spatial calibration results, and time stamp results are written into the multi-source planning data, and then linked and organized according to data source, regional affiliation, spatial location, and time stamp to generate the original dataset for charging facility planning. In the specific implementation process, the target area is delineated according to the administrative region boundary, the power supply grid boundary, and the traffic service area boundary, with the area of ​​a single target area set to 5km² to 50km²; the time granularity of the charging demand forecast data is set to 15 minutes, and the spatial granularity is set to a 250m×250m or 500m×500m grid; vehicle travel data, parking resource data, and existing charging facility operation data are accessed from the traffic monitoring platform, parking management platform, and charging facility operation platform, respectively, with the parking resource data update cycle set to 1 hour and the existing charging facility operation data collection cycle set to 15 minutes. The distribution network basic data and distribution network operation data are accessed from the distribution automation system. The collection cycle for distribution transformer load rate, feeder load rate, voltage deviation, and line current carrying capacity is set to 5 to 15 minutes. The unified coding adopts a 24-bit coding structure, the spatial location calibration adopts a unified coordinate reference, and the location deviation is controlled within 5 meters. The time reference adopts a unified standard time format, and the time deviation of operation data is controlled within 30 seconds and the time deviation of event data is controlled within 5 minutes. After the unified identification results, spatial calibration results, and time stamp results are written, the original dataset for charging facility planning is generated.

[0020] In this embodiment, S2 specifically refers to: Read the unified identification results, spatial calibration results, time stamp results, and planning association fields from the original dataset of charging facility planning to generate a planning dataset to be processed; Perform field integrity checks, duplicate record merging, abnormal record filtering, and format standardization on the planning dataset to be processed to generate a valid planning dataset; Based on a unified data format and a unified time granularity, the effective planning dataset is standardized and transformed, and spatiotemporally aligned according to the unified identification results, spatial calibration results, and time stamp results to generate a standardized planning dataset. Charging demand features are extracted from the charging demand forecast data and vehicle travel data in the standardized planning dataset, and construction resource features are extracted from the parking resource data and construction resource data in the standardized planning dataset. Existing facility features are extracted from existing charging facility operation data in the standardized planning dataset, and distribution network constraint features are extracted from distribution network basic data, distribution network operation data, and planning and construction data in the standardized planning dataset. The characteristics of charging demand, construction resources, existing facilities, and distribution network constraints are normalized and correlated to generate a set of charging facility planning characteristics. In the specific implementation process, when verifying field integrity, the unified identifier result, spatial calibration result, and time stamp result are mandatory fields. Records with a missing rate of more than 10% for key fields are entered into the supplementary entry queue. Duplicate records are merged according to the unified identifier result and time stamp result. Records of the same type with an interval of less than 30 seconds and a spatial distance of less than 5 meters are treated as the same record. When screening abnormal records, records with negative charging demand forecast values, fewer than 0 parking spaces, existing charging facilities operating power exceeding 120% of rated power, or transformer load rate exceeding 150% are marked as abnormal records. When processing the format uniformly, the spatial coordinates are uniformly set to the CGCS2000 coordinate system, and the time format is uniformly set to year-month-day-hour-minute-second.

[0021] During standardization, the time granularity of charging demand forecast data and existing charging facility operation data is unified to 15 minutes, while the time granularity of distribution network operation data is aligned from 5 minutes to 15 minutes. The spatial granularity is unified to a 250m×250m or 500m×500m grid. Charging demand characteristics include predicted charging frequency, predicted charging volume, and peak-hour demand intensity. Construction resource characteristics include the number of parking spaces, available construction area, and spatial accessibility. Existing facility characteristics include the number of existing charging piles, average utilization rate, and service saturation. Distribution network constraint characteristics include transformer capacity margin, feeder load rate, voltage deviation, and line current carrying capacity. Normalization is performed using a 0 to 1 interval normalization method. After correlation and processing, a charging facility planning feature set is generated.

[0022] In this embodiment, S3 specifically refers to: Read the charging demand characteristics, construction resource characteristics, existing facility characteristics, and distribution network constraint characteristics from the charging facility planning feature set, and generate site screening input data; Based on the characteristics of construction resources, spatial units within the target area that can be used for the construction of electric vehicle charging facilities are identified. Based on land use status, parking capacity, construction restrictions and spatial accessibility, the spatial units are initially screened to generate an initial set of constructible spaces. Based on the characteristics of charging demand, determine the charging demand intensity and service coverage corresponding to the initial buildable space set, and remove space units whose charging demand intensity does not meet the preset planning requirements to generate a demand-adaptive space set. Based on the characteristics of existing facilities, determine the service overlap between the demand-adaptive space set and existing charging facilities, and eliminate space units whose service overlap exceeds a preset threshold and do not have the necessity for new construction, thereby generating a candidate construction space set. Based on the distribution network constraint characteristics, the candidate construction space set is screened according to the basic access conditions, and the candidate construction space that meets the basic access conditions is determined as the candidate charging facility site, thus generating the candidate charging facility site set. In the specific implementation process, the spatial units available for the construction of electric vehicle charging facilities within the target area are divided into 250m×250m or 500m×500m grids, and are screened in conjunction with parking lots, public building ancillary land, commercial service land, and open land around transportation hubs; spatial units whose land use status is prohibited construction land, ecological protection land, or road red line occupied area are directly eliminated; spatial units with fewer than 20 parking spaces, a buildable area of ​​less than 80m², or a distance of more than 800m from the main road are not included in the initial set of buildable spaces; spatial accessibility is determined according to walking service distance and vehicle access distance, with walking service distance controlled between 300m and 800m and vehicle access distance controlled between 1km and 3km.

[0023] Based on the characteristics of charging demand, the charging demand intensity of each spatial unit is calculated. Spatial units with a predicted average daily charging frequency of less than 20 times or peak-hour demand intensity below a preset threshold are eliminated. Service overlap is assessed for spatial units within the service coverage of existing charging facilities. When the service overlap exceeds 70% and the average utilization rate of existing charging facilities is less than 60%, it is determined that there is no need for new construction. When screening basic access conditions, the access distance between candidate construction spaces and distribution network access nodes is preferably controlled within 500m, with a maximum of 800m. The distribution transformer capacity margin is not less than 20%, the feeder load rate is not higher than 85%, and the voltage deviation is controlled within ±7%. Candidate construction spaces that meet the above conditions are determined as candidate charging facility sites, and a set of candidate charging facility sites is generated.

[0024] In this embodiment, S4 specifically refers to: Read the set of candidate charging facility sites and extract charging demand features and distribution network constraint features from the charging facility planning feature set; Based on the characteristics of charging demand, determine the intensity, service range and time period of charging demand corresponding to different spatial units in the target area, and generate charging demand distribution information. Based on the constraints of the distribution network, candidate access nodes, power supply paths, capacity margins, voltage margins, and line load margins are determined to generate distribution network access conditions. Using the candidate charging facility sites in the candidate charging facility site set as site nodes, the demand spatial units corresponding to the charging demand distribution information as demand nodes, and the candidate access nodes corresponding to the distribution network access conditions as grid nodes, a planning associated node set is constructed. Establish site-demand correlation based on the spatial coverage and demand matching relationships between candidate charging facility sites and demand spatial units; Establish site access associations based on the matching status of access distance, power supply path, and access capacity between candidate charging facility sites and candidate access nodes; Based on the capacity margin, voltage margin, and line load margin corresponding to the candidate access nodes, establish the distribution network constraint relationship; The planning-related node set, site demand relationship, site access relationship and distribution network constraint relationship are coupled and modeled to construct a planning constraint relationship model that couples site demand and distribution network constraints. Based on the planning constraint association model, the demand coverage results, access matching results and distribution network constraint matching results corresponding to the candidate charging facility sites are output, and the planning constraint association results are generated. In the specific implementation process, charging demand distribution information is generated according to 250m×250m or 500m×500m spatial units, and the demand time period distribution is statistically analyzed according to 15-minute time granularity. The average daily charging frequency, peak demand intensity, and predicted charging volume are written into the demand node. Candidate access nodes are selected from distribution transformer low-voltage side access points, 10kV line branch points, and planned reserved access points within 800m of the candidate charging facility site. The capacity margin is not less than 20%, the voltage margin corresponds to the node voltage deviation controlled within ±7%, and the line load margin corresponds to the feeder load rate not exceeding 85%. Site demand correlation is established based on service radius and demand intensity. The service radius of public slow charging sites is set to 500m to 800m, and the service radius of fast charging sites is set to 1km to 3km. Site access correlation is established based on access distance, power supply path level, and access capacity matching status. No valid correlation is established when the access distance exceeds 800m or the power supply path is not connected. Distribution network constraint correlation is set with constraint labels based on capacity margin, voltage margin, and line load margin. Site nodes, demand nodes, and grid nodes are indexed and bound using unified site number, spatial unit number, and access node number to generate a planning constraint correlation model. In the planning constraint correlation results, demand coverage results are represented by covered demand and uncovered demand, access matching results are represented by accessible, capacity-limited access, and inaccessible access, and distribution network constraint matching results are represented by satisfied constraints, critical constraints, and unsatisfied constraints.

[0025] In this embodiment, the planning-related node set, site demand relationship, site access relationship, and distribution network constraint relationship are coupled and modeled to construct a planning-constraint relationship model that couples site demand and distribution network constraints. Specifically, the model is as follows: Label the site nodes, demand nodes, and power grid nodes in the planning associated node set with node type annotations, and generate planning node type annotation results; The relationship types of site demand, site access, and distribution network constraint are labeled, and the planning relationship type labeling results are generated. Based on the planning node type labeling results and the planning relationship type labeling results, a three-layer planning association structure including the site layer, demand layer and power grid layer is constructed. Based on the site demand relationship, the site nodes in the site layer are connected with the demand nodes in the demand layer to generate a site demand coupled substructure; Based on the site access association, the site nodes in the site layer are connected to the power grid nodes in the power grid layer to generate a site access coupling substructure; Based on the distribution network constraint relationships, the grid nodes in the grid layer are constrained and bound to their corresponding capacity margin, voltage margin, and line load margin to generate a grid constraint coupling substructure. The site demand coupling substructure, the site access coupling substructure, and the power grid constraint coupling substructure are indexed in a unified manner to generate the site demand access constraint coupling structure. Based on the site demand access constraint coupling structure, the demand coverage status, access matching status and distribution network constraint status corresponding to each candidate charging facility site are determined, and a planning constraint association model coupling site demand and distribution network constraint is constructed. In the specific implementation process, site nodes, demand nodes, and grid nodes are labeled with different node type identifiers. Site node fields include site number, buildable area, parking capacity, and candidate facility type. Demand node fields include spatial unit number, average daily charging frequency, peak demand intensity, and predicted charging volume. Grid node fields include access node number, distribution transformer capacity margin, voltage margin, and line load margin. Site-demand relationships are marked as first-type relationships, site-access relationships as second-type relationships, and distribution network constraint relationships as third-type relationships. Each relationship records the starting node, ending node, relationship type, and matching status. The three-layer planning association structure is organized according to the "demand layer - site layer - power grid layer". In the site demand coupling substructure, demand nodes within the service radius are connected to site nodes, and the connection status is divided into high coverage, medium coverage and low coverage according to the demand coverage ratio. In the site access coupling substructure, candidate charging facility sites are connected to candidate access nodes within 800m. If the access distance exceeds 800m or the power supply path is not connected, no effective connection is established. In the power grid constraint coupling substructure, power grid nodes with a capacity margin of less than 20%, a feeder load rate of more than 85%, or a voltage deviation of more than ±7% are marked as constrained and unmet. After the unified index is completed, the demand coverage status, access matching status and distribution network constraint status are bound to the same candidate charging facility site according to the site number, and a planning constraint association model coupling site demand and distribution network constraints is constructed.

[0026] In this embodiment, S5 specifically refers to: Read the demand coverage results, access matching results, and distribution network constraint matching results from the planning constraint association results, and generate access feasibility verification input data; Based on the input data for access feasibility verification, the candidate access nodes, access capacity and access distance corresponding to each candidate charging facility site are determined, and access capability verification results are generated. Based on the access matching results, determine the power supply lines, upstream distribution transformers, associated feeders and power supply levels between each candidate charging facility site and the candidate access node, and generate power supply path verification results; Based on the distribution network constraint matching results, determine the capacity margin, voltage margin, line load margin, and power supply reliability status of each candidate access node, and generate the operation safety boundary verification results; Based on the access capacity verification results, power supply path verification results, and operation safety boundary verification results, candidate charging facility sites that do not meet the requirements of access capacity, access distance, power supply path connectivity, and operation safety boundary are eliminated, and a set of feasible candidate sites is generated. Based on the set of feasible candidate sites, determine the accessible nodes, constructable capacity range, optional charging facility types and access constraints corresponding to each feasible candidate site, and generate candidate charging facility planning schemes; The candidate charging facility planning schemes corresponding to each feasible candidate site are summarized to generate a set of candidate charging facility planning schemes. In the specific implementation process, the access capacity verification is carried out on a per-candidate charging facility site basis. Candidate access nodes are preferentially selected from distribution transformer low-voltage side access points and 10kV line branch points within 500m of the candidate charging facility site, with a maximum access distance of no more than 800m. The available access capacity is determined according to the remaining capacity of the candidate access nodes, with a distribution transformer capacity margin of no less than 20%, and the planned access capacity of a single candidate charging facility site does not exceed 80% of the available capacity of the corresponding candidate access node. The power supply path verification is performed in the following order: "candidate charging facility site - candidate access node - upstream distribution transformer - associated feeder". The power supply level shall not exceed 3 levels. A power supply path verification result is generated when there are no breaks, reverse power supply or topology errors in the power supply line. When verifying the operation safety boundary, the feeder load rate shall not exceed 85%, the node voltage deviation shall be controlled within ±7%, the line current carrying capacity shall not exceed 90%, and the power supply reliability status shall not be lower than the preset reliability level. Candidate charging facility sites that do not meet the above conditions shall be removed from the candidate range. The remaining sites shall generate a set of candidate charging facility planning schemes according to the access nodes, the range of buildable capacity, the optional charging facility types and the access constraint status.

[0027] In this embodiment, S6 specifically refers to: Read the feasible candidate sites, accessible nodes, constructable capacity range, optional charging facility types and access constraint status from the candidate charging facility planning scheme set, and generate planning optimization input data; Based on the planning optimization input data, determine the construction location variables, construction scale variables, facility type variables, access node variables, and distribution network transformation method variables, and generate an optimization decision variable set; Based on the charging demand characteristics, construction resource characteristics, existing facility characteristics, and distribution network constraint characteristics in the planning optimization input data, a set of planning optimization objectives is generated; Generate a set of distribution network constraints based on the access constraint status, accessible nodes, and available capacity range; The set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints are input into the improved MISOCP model. By improving the MISOCP model, the construction location, construction scale, facility type, access node and distribution network transformation method are jointly solved to generate a target planning scheme that meets the distribution network constraints. Based on the target planning scheme, determine the construction status, construction capacity, charging facility configuration results, access node selection results and power grid transformation methods for each feasible candidate site, and generate charging facility planning optimization results; In the specific implementation process, the planning optimization input data is input in batches according to a single target area. The number of feasible candidate sites for each input is set to 20 to 200, and the number of accessible nodes is set to 10 to 100. The construction location variable uses 0 or 1 to indicate whether to construct. The construction scale variable is set to a continuous value range of 50kW to 1200kW. The facility type variable is coded according to three categories: slow charging, fast charging, and supercharging. The access node variable is discretely valued according to the candidate access node number. The distribution network transformation method variable is coded according to no transformation, transformer capacity expansion, line transformation, and access point expansion. In the set of optimization targets, the charging demand coverage rate is no less than 85%, and areas with existing facility service saturation exceeding 80% are prioritized for optimization. In the set of distribution network constraints, the transformer load rate is controlled within 80%, the feeder load rate is controlled within 85%, the node voltage deviation is controlled within ±7%, and the line current carrying capacity occupancy rate is controlled within 90%. When solving the improved MISOCP model, the maximum number of iterations is set to 500, the optimal gap threshold is set to 0.01, and after the solution is completed, the target planning scheme that meets the distribution network constraints is output, and the corresponding construction status, construction capacity, charging facility configuration results, access node selection results, and distribution network transformation methods are generated.

[0028] In this embodiment, the set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints are specifically input into the improved MISOCP model as follows: The improved MISOCP model includes a planning input module, a variable encoding module, an objective construction module, a second-order cone constraint transformation module, a feasible region shrinkage module, a joint solution module, and a result output module. The planning input module reads the set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints, and extracts the construction location variables, construction scale variables, facility type variables, access node variables, distribution network transformation method variables, planning optimization objectives, and distribution network operation constraints. The variable coding module encodes the construction location, facility type, access node, and distribution network renovation methods as integer variables, and the construction scale as a continuous variable, generating a mixed integer decision variable set. The objective construction module constructs a multi-objective planning function set based on the planning optimization objective set, including demand coverage objective, construction cost objective, access and transformation cost objective, distribution network safety margin objective, and facility utilization efficiency objective; The voltage safety constraint, line capacity constraint, and transformer capacity constraint in the distribution network constraint set are transformed into second-order cone constraints through the second-order cone constraint transformation module. The access capacity constraint, power supply path constraint, and access distance constraint are embedded as mixed integer constraints to generate second-order cone constraint conditions. The feasible region shrinking module shrinks the feasible region based on the access constraint status in the candidate charging facility planning scheme set, and shrinks the construction location variables, access node variables and construction scale variables that do not meet the access feasibility verification requirements to generate a shrunken planning feasible region. The joint solution module solves the mixed integer decision variable set, the multi-objective programming function set, the second-order cone constraint condition, and the shrinking programming feasible region together to generate the optimization solution result. The results output module outputs values ​​for construction location, construction scale, facility type, access node, and distribution network transformation method based on the optimization solution results. In its implementation, the improved MISOCP model inherits the framework of the traditional mixed-integer second-order cone programming model for jointly solving discrete programming variables and continuous capacity variables, including variable modeling, constraint transformation, and optimal solution processes. However, it introduces a feasible region shrinkage module and a distribution network operation constraint embedding mechanism to replace the traditional MISOCP model's method of directly solving all candidate schemes uniformly. First, the construction location variable, facility type variable, access node variable, and distribution network renovation method variable are set as integer variables. The construction location variable takes values ​​of 0 or 1, the facility type variable is coded into three categories: slow charging, fast charging, and supercharging, and the distribution network renovation method variable is coded into four categories: no renovation, transformer capacity expansion, line renovation, and access point expansion. The construction scale variable is set as a continuous variable, with a value range of 50kW to 1200kW. Subsequently, the voltage safety constraint, line capacity constraint, and transformer capacity constraint are transformed through the second-order cone constraint transformation module, and the access distance, power supply path connectivity, and access capacity limit are embedded as mixed-integer constraints in the model. Compared with the traditional MISOCP model, this invention eliminates unaccessible schemes based on access constraint status before solving, and narrows the range of construction scale values ​​for candidate access nodes with insufficient capacity margin. This reduces the invalid search space, improves the efficiency of planning and solving, and ensures that the output construction location, construction scale, facility type, access node, and distribution network transformation method meet the charging service requirements and the safety operation boundary of the distribution network.

[0029] In this embodiment, S7 specifically refers to: Read the values ​​of construction location, construction scale, facility type, access node, and distribution network transformation method from the charging facility planning optimization results, and generate planning decision input data; Based on the values ​​for construction location and construction scale, the recommended construction locations and recommended construction capacity of electric vehicle charging facilities within the target area are determined, and a construction plan is generated. Based on the facility type value, determine the charging facility type, facility configuration quantity and service target range corresponding to each recommended construction location, and generate facility configuration results; Based on the access node values, determine the access power point, access line, and power supply path corresponding to each recommended construction location, and generate an access plan; Based on the value of the distribution network renovation method, determine the corresponding renovation content for transformer capacity expansion, line renovation, access point expansion and protection configuration adjustment, and generate distribution network renovation requirements. Based on the construction plan, access plan, facility configuration results and power distribution network transformation needs, the recommended construction locations within the target area are sorted in order of implementation to generate implementation priorities. The construction plan, access plan, facility configuration results, power distribution network transformation needs and implementation priorities are combined to generate planning and decision-making information for electric vehicle charging facilities. In the specific implementation process, it is recommended that the construction location be output according to the target area, power grid and site number, and the construction capacity be graded from 50kW to 1200kW; the charging facility type should be configured according to slow charging, fast charging and supercharging, with the power of a single slow charging pile set to 7kW to 14kW, the power of a single fast charging pile set to 60kW to 180kW, and the power of a single supercharging pile set to 250kW to 480kW; In the access scheme, the access distance between the access power point and the recommended construction location is controlled within 800m. The access lines and power supply paths are recorded step by step according to the candidate access nodes, the upstream distribution transformers, and the associated feeders. The distribution network transformation needs are output in four categories: distribution transformer capacity expansion, line transformation, access point expansion, and protection configuration adjustment. The implementation priority is ranked according to the charging demand coverage rate, the distribution network constraint satisfaction status, and the construction and transformation costs. The ranking results are divided into three levels: priority implementation, near-term implementation, and long-term reserve. Finally, the planning decision information for electric vehicle charging facilities is generated.

[0030] Example 1: To verify the feasibility of this invention in practice, it was applied to the planning and decision-making scenario of electric vehicle charging facilities in the eastern new district of a city. This area is a mixed distribution of newly built residential areas, industrial parks, commercial complexes, and transportation hubs, covering an area of ​​approximately 32.6 km². There are 46 existing parking lots with approximately 18,520 public parking spaces and 428 existing public charging piles, including 276 slow charging piles and 152 fast charging piles. The number of new energy vehicles in this area has increased rapidly in the past year. During morning and evening rush hours, commuter vehicles concentrate in the vicinity of industrial parks and subway transfer stations, and the demand for parking and charging in residential areas increases significantly at night. The existing charging facilities suffer from spatial distribution problems such as overlapping locations in commercial centers, insufficient supply in residential areas, and long waiting times in some industrial parks. Meanwhile, the power distribution network in this area is supplied by three 110kV substations and 18 10kV feeders. Some feeders have a load rate of nearly 80% during the summer evening peak, and some old distribution areas have insufficient capacity to connect to the network. If charging facilities are only deployed according to the intensity of charging demand, it is easy to cause problems such as transformer overload, increased feeder load, and increased node voltage deviation.

[0031] In this scenario, data on charging demand forecasts, vehicle travel, parking resources, construction resources, existing charging facility operation, power distribution network infrastructure, power distribution network operation, and planned construction were collected over three consecutive months within the region. Charging demand forecasts were statistically analyzed at a 15-minute time granularity. Vehicle travel data included the number of vehicles entering and leaving the area, their dwell time, and destination type. Parking resource data included the total number of parking spaces, average turnover rate, and peak-hour available parking spaces. Existing charging facility operation data included charging frequency, charging volume, average utilization rate, and queuing records. Power distribution network operation data included transformer load rate, feeder load rate, voltage deviation, and line current carrying capacity. After unified coding, spatial calibration, and time alignment, approximately 1.864 million valid planning data entries were generated, covering 112 250m×250m spatial units, 46 parking resource points, 428 existing charging piles, 312 transformer nodes, and 18 10kV feeders.

[0032] When applying this invention, the first step is to screen the buildable spaces within the target area based on the charging facility planning feature set. After preliminary screening based on land use status, parking capacity, construction restrictions, and spatial accessibility, 68 initial buildable spaces are obtained from 112 spatial units. Further, by combining the charging demand intensity and the service overlap of existing charging facilities, areas with sufficient existing facility coverage and an average utilization rate of less than 60% are eliminated, resulting in 49 candidate construction spaces. Then, by combining the distribution network constraint characteristics, basic access conditions are screened, eliminating spatial units with transformer capacity margins of less than 20%, feeder load rates of more than 85%, and access distances exceeding 800m, ultimately generating a set of 37 candidate charging facility sites.

[0033] Subsequently, a planning constraint correlation model coupling site demand and distribution network constraints was constructed for 37 candidate charging facility sites. This model treats candidate charging facility sites as site nodes, charging demands corresponding to different spatial units as demand nodes, and accessible distribution transformer nodes and 10kV line branch points as grid nodes, establishing site demand correlation, site access correlation, and distribution network constraint correlation, respectively. The model clearly shows that while some commercial complexes have high charging demand, nearby distribution transformer capacity margins are insufficient, and directly adding fast charging piles would cause overload exceeding limits; some residential areas have high charging demand and large off-peak load space at night, making slow charging piles suitable for deployment; industrial park entrances and exits have strong daytime fast charging demand, but this needs to be planned in conjunction with feeder line renovation schemes.

[0034] After completing the feasibility verification for access, a total of 54 candidate charging facility planning schemes were formed. These candidate schemes were then input into the improved MISOCP model, which jointly optimized the construction location, scale, facility type, access nodes, and distribution network modification methods. During optimization, the construction scale range was set to 50kW to 1200kW, with slow charging piles configured with a single pile power of 7kW to 14kW, fast charging piles with a single pile power of 60kW to 180kW, and supercharging piles with a single pile power of 250kW to 480kW. In terms of distribution network constraints, the transformer load rate was controlled within 80%, the feeder load rate within 85%, the node voltage deviation within ±7%, and the line current carrying capacity occupancy rate within 90%. After optimization, this invention recommends the construction of 21 charging facility sites, including 9 slow charging sites in residential areas, 5 fast charging sites in commercial areas, 5 mixed fast and supercharging sites in industrial parks, and 2 fast charging sites at transportation hubs. The recommended total number of new charging piles is 316, including 198 slow charging piles, 104 fast charging piles, and 14 supercharging piles, with a total new capacity of approximately 18.72MW. Simultaneously, it also recommends the capacity expansion of 6 distribution transformers, the upgrading of 8 low-voltage lines, the expansion of 5 access points, and the adjustment of 4 protection configurations.

[0035] To further verify the beneficial effects of this invention, it was compared with traditional empirical planning methods and ordinary MISOCP planning methods under the same data range, the same candidate sites, and the same distribution network operation boundary. Traditional empirical planning methods mainly determine construction locations based on charging demand and parking resource scale, without fully embedding distribution network operation constraints; ordinary MISOCP planning methods can perform site selection and capacity optimization, but do not combine planning constraint correlation models and access feasibility verification results to shrink the feasible region. The comparison results are shown in Table 1.

[0036] Table 1 Comparison of the effects of charging facility planning decisions As shown in Table 1, the present invention outperforms the comparative methods in terms of charging demand coverage, distribution network safety constraint satisfaction, access upgrade costs, and planning efficiency. While traditional empirical planning methods can quickly determine construction locations, they tend to concentrate charging facilities in commercial centers and traffic hotspots, leading to some distribution transformer nodes exceeding their limits and insufficient coverage of nighttime slow charging demand in residential areas. The ordinary MISOCP planning method can reduce some construction redundancy, but due to the lack of a planning constraint correlation model that couples site requirements with distribution network constraints, some candidate solutions still need to be repeatedly eliminated during the solution process, resulting in a long solution time. The present invention expresses the coupling relationship between candidate sites, charging demand, and distribution network access conditions in advance through a planning constraint correlation model, and narrows the range of candidate solutions through access feasibility checks. This allows the improved MISOCP model to complete joint optimization within a smaller feasible domain, improving planning efficiency and preventing construction schemes that do not meet distribution network safety boundaries from entering the final result.

[0037] The implementation results show that this invention can incorporate the carrying capacity of the distribution network, power supply path, and operational safety boundary into the planning and decision-making process of charging facilities while meeting the demand for charging services. This solves the problems of traditional charging facility planning, which emphasizes demand over access, quantity over constraints, and the disconnect between construction plans and distribution network upgrade plans. The electric vehicle charging facility planning and decision-making information generated by this invention not only includes recommended construction locations and capacity, but also simultaneously provides access power sources, access lines, distribution network upgrade requirements, and implementation priorities. This information can be directly used for regional charging facility construction planning and the formulation of supporting distribution network upgrade plans.

[0038] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A planning and decision-making method for electric vehicle charging facilities based on distribution network constraints, characterized in that, Includes the following steps: S1. Collect multi-source planning data related to electric vehicle charging facility planning within the target area, process the multi-source planning data in a unified manner, and generate the original dataset of charging facility planning. S2. Standardize the original dataset of the charging facility planning and extract charging demand features, construction resource features, existing facility features and distribution network constraint features to generate a charging facility planning feature set. S3. Based on the charging facility planning feature set, the available space within the target area is screened to generate a set of candidate charging facility sites; S4. Based on the candidate charging facility site set and the charging facility planning feature set, determine the charging demand distribution information and distribution network access conditions, and construct a planning constraint association model that couples site demand with distribution network constraints, and generate planning constraint association results. S5. Based on the planning constraint association results, verify the access feasibility of the candidate charging facility sites in terms of access capacity, power supply path and operation safety boundary, and generate a set of candidate charging facility planning schemes. S6. Input the set of candidate charging facility planning schemes into the improved MISOCP model, and use the improved MISOCP model to jointly optimize the construction location, construction scale, facility type, access node and distribution network transformation method to generate charging facility planning optimization results; S7. Based on the optimization results of the charging facility planning, determine the construction plan, access plan, power distribution network transformation requirements and implementation priority of the electric vehicle charging facility, and generate electric vehicle charging facility planning decision information.

2. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 1, characterized in that, Specifically, S1 is: Acquire charging demand forecast data, vehicle travel data, parking resource data, construction resource data, existing charging facility operation data, power distribution network basic data, power distribution network operation data, and planning and construction data within the target area to generate multi-source planning data; Read the regional affiliation information, site affiliation information, facility affiliation information, equipment affiliation information, and power supply affiliation information from the multi-source planning data, and encode them uniformly according to the preset coding rules to generate a unified identification result; Based on the location description information, spatial coordinate information, power supply range information, and distribution network access location information in the multi-source planning data, the spatial location of the charging demand occurrence area, parking resource location, construction resource location, existing charging facility location, and distribution network access node location is spatially calibrated to generate spatial calibration results. Based on the demand forecast time, vehicle travel time, parking record time, charging facility operation time, power distribution network operation data collection time, and planning and construction time in the multi-source planning data, a unified time benchmark is generated to produce a time stamp result. The unified identification result, the spatial calibration result, and the time stamp result are written into the multi-source planning data, and then linked and organized according to data source, regional affiliation, spatial location, and time stamp to generate the original dataset for charging facility planning.

3. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 1, characterized in that, Specifically, S2 is: Read the unified identification results, spatial calibration results, time stamp results, and planning association fields from the original dataset of the charging facility planning, and generate a planning dataset to be processed; The planning dataset to be processed is subjected to field integrity verification, duplicate record merging, abnormal record filtering and format unification processing to generate a valid planning dataset; Based on a unified data format and a unified time granularity, the effective planning dataset is standardized and transformed, and spatiotemporally aligned according to the unified identification results, the spatial calibration results, and the time stamp results to generate a standardized planning dataset. Charging demand features are extracted from the charging demand forecast data and vehicle travel data in the standardized planning dataset, and construction resource features are extracted from the parking resource data and construction resource data in the standardized planning dataset. Existing facility features are extracted based on existing charging facility operation data in the standardized planning dataset, and distribution network constraint features are extracted based on distribution network basic data, distribution network operation data, and planning and construction data in the standardized planning dataset. The charging demand characteristics, construction resource characteristics, existing facility characteristics, and distribution network constraint characteristics are normalized and correlated to generate a charging facility planning feature set.

4. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 1, characterized in that, Specifically, S3 is: Read the charging demand characteristics, construction resource characteristics, existing facility characteristics and distribution network constraint characteristics from the charging facility planning feature set, and generate site screening input data; Based on the aforementioned construction resource characteristics, spatial units within the target area that can be used for the construction of electric vehicle charging facilities are determined, and these spatial units are initially screened based on land use status, parking capacity, construction restrictions, and spatial accessibility to generate an initial set of constructable spaces. Based on the charging demand characteristics, determine the charging demand intensity and service coverage corresponding to the initial buildable space set, and remove space units whose charging demand intensity does not meet the preset planning requirements to generate a demand-adaptive space set. Based on the characteristics of the existing facilities, the service overlap status between the demand-adaptive space set and the existing charging facilities is determined, and space units with service overlap exceeding a preset threshold and no need for new construction are eliminated, thereby generating a candidate construction space set. Based on the distribution network constraint characteristics, the candidate construction space set is screened for basic access conditions, and the candidate construction spaces that meet the basic access conditions are determined as candidate charging facility sites, thus generating a candidate charging facility site set.

5. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 1, characterized in that, Specifically, S4 is: Read the candidate charging facility site set and extract charging demand features and power distribution network constraint features from the charging facility planning feature set; Based on the charging demand characteristics, determine the charging demand intensity, demand service range and demand time distribution corresponding to different spatial units within the target area, and generate charging demand distribution information. Based on the aforementioned distribution network constraint characteristics, candidate access nodes, power supply paths, capacity margins, voltage margins, and line load margins are determined to generate distribution network access conditions. Using the candidate charging facility sites in the candidate charging facility site set as site nodes, the demand spatial units corresponding to the charging demand distribution information as demand nodes, and the candidate access nodes corresponding to the distribution network access conditions as grid nodes, a planning associated node set is constructed. Based on the spatial coverage and demand matching relationships between the candidate charging facility sites and the demand spatial units, a site demand correlation relationship is established. Based on the matching status of the access distance, power supply path, and access capacity between the candidate charging facility sites and the candidate access nodes, a site access association relationship is established. Based on the capacity margin, voltage margin, and line load margin corresponding to the candidate access nodes, establish the distribution network constraint relationship; The planning-related node set, the site demand relationship, the site access relationship, and the distribution network constraint relationship are coupled and modeled to construct a planning constraint relationship model that couples site demand and distribution network constraints. Based on the planning constraint association model, the demand coverage result, access matching result and distribution network constraint matching result corresponding to the candidate charging facility sites are output, and the planning constraint association result is generated.

6. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 5, characterized in that, The specific steps for coupling and modeling the planning-related node set, the site demand relationship, the site access relationship, and the distribution network constraint relationship to construct a planning-constraint relationship model that couples site demand and distribution network constraints are as follows: The site nodes, demand nodes, and power grid nodes in the planning associated node set are labeled with node types to generate planning node type labeling results; The relationship types of the site demand relationship, the site access relationship, and the distribution network constraint relationship are labeled to generate planning relationship type labeling results; Based on the planning node type labeling results and the planning relationship type labeling results, a three-layer planning association structure including the site layer, demand layer and power grid layer is constructed. Based on the site demand association, the site nodes in the site layer are connected to the demand nodes in the demand layer to generate a site demand coupled substructure; Based on the site access association, the site nodes in the site layer are connected to the power grid nodes in the power grid layer to generate a site access coupling substructure; Based on the distribution network constraint association, the grid nodes in the grid layer are constrained and bound to the corresponding capacity margin, voltage margin and line load margin to generate a grid constraint coupling substructure; The site demand coupling substructure, the site access coupling substructure, and the power grid constraint coupling substructure are indexed in a unified manner to generate a site demand access constraint coupling structure. Based on the site demand access constraint coupling structure, the demand coverage status, access matching status and distribution network constraint status corresponding to each candidate charging facility site are determined, and a planning constraint association model coupling site demand and distribution network constraint is constructed.

7. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 1, characterized in that, Specifically, S5 is: Read the demand coverage result, access matching result and distribution network constraint matching result from the planning constraint association result, and generate access feasibility verification input data; Based on the access feasibility verification input data, determine the candidate access nodes, access capacity and access distance corresponding to each candidate charging facility site, and generate access capability verification results; Based on the access matching results, determine the power supply lines, upstream distribution transformers, associated feeders and power supply levels between each candidate charging facility site and the candidate access node, and generate power supply path verification results; Based on the distribution network constraint matching results, determine the capacity margin, voltage margin, line load margin, and power supply reliability status of each candidate access node, and generate the operation safety boundary verification results; Based on the access capacity verification results, the power supply path verification results, and the operation safety boundary verification results, candidate charging facility sites that do not meet the requirements of access capacity, access distance, power supply path connectivity, and operation safety boundary are eliminated to generate a set of feasible candidate sites. Based on the set of feasible candidate sites, determine the accessible nodes, constructable capacity range, optional charging facility types and access constraint status corresponding to each feasible candidate site, and generate a candidate charging facility planning scheme; The candidate charging facility planning schemes corresponding to each feasible candidate site are summarized to generate a set of candidate charging facility planning schemes.

8. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 1, characterized in that, Specifically, S6 is: Read the feasible candidate sites, accessible nodes, constructible capacity range, optional charging facility types, and access constraint status from the candidate charging facility planning scheme set, and generate planning optimization input data; Based on the planning optimization input data, determine the construction location variables, construction scale variables, facility type variables, access node variables, and distribution network transformation method variables, and generate an optimization decision variable set; Based on the charging demand characteristics, construction resource characteristics, existing facility characteristics, and distribution network constraint characteristics in the planning optimization input data, a set of planning optimization objectives is generated; Based on the access constraint status, the accessible nodes, and the buildable capacity range, a set of distribution network constraint conditions is generated. Input the set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints into the improved MISOCP model; The improved MISOCP model is used to jointly solve for the construction location, construction scale, facility type, access node and distribution network transformation method to generate a target planning scheme that meets the distribution network constraints. Based on the target planning scheme, determine the construction status, construction capacity, charging facility configuration results, access node selection results, and power distribution network transformation methods for each feasible candidate site, and generate charging facility planning optimization results.

9. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 8, characterized in that, The specific steps of inputting the set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints into the improved MISOCP model are as follows: The improved MISOCP model includes a planning input module, a variable encoding module, an objective construction module, a second-order cone constraint transformation module, a feasible region shrinkage module, a joint solution module, and a result output module. The planning input module reads the set of optimization decision variables, the set of planning optimization objectives, and the set of distribution network constraints, and extracts construction location variables, construction scale variables, facility type variables, access node variables, distribution network transformation method variables, planning optimization objectives, and distribution network operation constraints. The variable encoding module encodes the construction location variable, facility type variable, access node variable, and distribution network renovation method variable as integer variables, and encodes the construction scale variable as a continuous variable, generating a mixed integer decision variable set. The target construction module constructs a multi-objective planning function set based on the planning optimization target set, which includes demand coverage target, construction cost target, access and transformation cost target, distribution network safety margin target, and facility utilization efficiency target; The second-order cone constraint transformation module performs second-order cone constraint transformation on the voltage safety constraint, line capacity constraint and distribution transformer capacity constraint in the distribution network constraint set, and embeds the access capacity constraint, power supply path constraint and access distance constraint as mixed integer constraints to generate second-order cone constraint conditions. The feasible region shrinking module shrinks the feasible region based on the access constraint status in the candidate charging facility planning scheme set, and shrinks the construction location variables, access node variables and construction scale variables that do not meet the access feasibility verification requirements to generate a shrunken planning feasible region. The joint solution module jointly solves the mixed integer decision variable set, the multi-objective programming function set, the second-order cone constraint condition, and the shrinking planning feasible region to generate an optimized solution result. The result output module outputs the values ​​for construction location, construction scale, facility type, access node, and power distribution network transformation method based on the optimization solution results.

10. The electric vehicle charging facility planning and decision-making method based on distribution network constraints according to claim 1, characterized in that, Specifically, S7 is: Read the values ​​of construction location, construction scale, facility type, access node, and power distribution network transformation method from the charging facility planning and optimization results, and generate planning decision input data; Based on the values ​​of the construction location and the construction scale, the recommended construction location and recommended construction capacity of electric vehicle charging facilities in the target area are determined, and a construction plan is generated. Based on the value of the facility type, determine the type of charging facility, the number of facilities configured, and the scope of service targets corresponding to each recommended construction location, and generate facility configuration results; Based on the values ​​of the access nodes, the access power point, access line and power supply path corresponding to each recommended construction location are determined, and an access scheme is generated. Based on the value of the power distribution network renovation method, determine the corresponding renovation content for transformer capacity expansion, line renovation, access point expansion and protection configuration adjustment, and generate power distribution network renovation requirements. Based on the construction plan, the access plan, the facility configuration results, and the power distribution network renovation requirements, the recommended construction locations within the target area are sorted in order of implementation to generate an implementation priority. The construction plan, the access plan, the facility configuration results, the power distribution network transformation requirements, and the implementation priority are combined to generate electric vehicle charging facility planning decision information.

Images