Big data-based city charging hotspot area prediction and planning method

Abstract

This invention relates to the field of prediction and optimization technology, specifically a method for predicting and planning urban charging hotspot areas based on big data. The method includes the following steps: acquiring traffic flow, electric vehicle paths, and charging layers; fusing them to generate spatial behavior data; calculating intersection density to filter hotspot areas; matching charging station distribution to identify areas with insufficient coverage; constructing deployment paths; matching trajectory extraction nodes; and outputting the charging hotspot prediction and planning results. In this invention, traffic flow, electric vehicle paths, and charging usage data are fused to construct a regional behavior model. Hotspot areas are identified by combining path intersection density, improving the positioning accuracy of high-demand areas. A coverage identification mechanism is introduced to enhance the targeting of site selection. Deployment paths are extracted based on boundary structures to enhance the continuity and scalability of the layout. Intersection nodes are extracted and encoded by combining time-period trajectories to achieve adaptation to dynamic changes in deployment, improving the rationality and practicality of the planning results.

Description

Technical Field

[0001] This invention relates to the field of prediction and optimization technology, and in particular to a method for predicting and planning urban charging hotspot areas based on big data. Background Technology

[0002] The field of prediction and optimization technology refers to the use of data analysis, mathematical models, and statistical methods to predict or optimize future trends, thereby assisting in decision-making and management. Data processing methods involved in this field include, but are not limited to, data mining, machine learning, and deep learning, aiming to extract patterns from large amounts of historical data for trend prediction, demand forecasting, and resource allocation optimization. Prediction and optimization technologies are widely used in various industries, including but not limited to finance, healthcare, energy, transportation, and urban planning. Through scientific prediction and optimization solutions, they help improve decision-making efficiency and resource allocation efficiency, reduce costs, and enhance the sustainable development capabilities of systems. Traditional big data-based methods for predicting and planning urban charging hotspots involve collecting and analyzing large amounts of charging demand data, traffic flow data, and geographic information in cities, using big data analysis and prediction technologies to identify and plan the location and quantity of charging piles. Traditional planning methods often rely on experience or relatively simple statistical analysis, lacking accurate prediction models and comprehensive planning systems, resulting in uneven distribution of charging facilities and an inability to effectively meet market demand. To address this issue, big data-based prediction and planning methods typically employ techniques such as data mining, predictive analysis, and spatial optimization. By combining historical and real-time data, they can predict charging hotspot areas, rationally plan the distribution of charging facilities, thereby improving the utilization efficiency of charging stations and reducing construction costs.

[0003] Existing technologies for predicting urban charging hotspots rely on static data aggregation, lacking a three-dimensional representation of charging behavior in different areas within a spatial range. They fail to fully reflect the interaction between traffic paths, electric vehicle trajectories, and facility distribution, resulting in accuracy deviations in hotspot location and an inability to effectively identify key areas with weak charging facility coverage. Furthermore, their path planning methods are mostly based on macro-regional layouts, ignoring the impact of urban boundary structures on the continuity of facility deployment. This makes it difficult to effectively adapt to dynamic trajectories, easily leading to node site selection deviating from actual traffic demands. This causes a disconnect between planning results and actual usage scenarios, affecting the practicality of facility deployment and the integrity of path connections, reducing resource allocation efficiency, and hindering the improvement of service capabilities. Summary of the Invention

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a method for predicting and planning urban charging hotspot areas based on big data, comprising the following steps: S1: Obtain the urban traffic flow layer, electric vehicle travel path layer and charging station usage distribution layer, construct a fusion table of the three types of layers based on the area number, and perform numbering integration on the overlapping positions of the layers according to the urban grid division rules to generate spatial charging behavior distribution data; S2: Based on the spatial charging behavior distribution data, calculate the aggregation density of electric vehicle path intersections in the urban grid, and generate hotspot traffic area data by filtering according to the aggregation density value; S3: Match the location of the hotspot traffic area data with the charging station distribution records within the traffic area, identify the area number with insufficient charging facility coverage, and output a list of urban hotspot area numbers. S4: Based on the map boundaries of the corresponding areas in the list of urban hotspot area numbers, extract the number of boundary path intersections and the spatial location of deployment candidates, construct continuous paths based on the geometric relationship of the connecting lines within the boundaries, and generate urban charging facility deployment path information; S5: Match the location of the urban charging facility deployment path information with the electric vehicle movement trajectory data of different time periods, extract the time period intersection nodes and encode them according to the path order, and output the prediction and planning results of urban charging hotspot areas.

[0005] As a further embodiment of the present invention, the three-layer fusion table includes a traffic flow number layer, an electric vehicle path number layer, and a charging station usage number layer. The spatial charging behavior distribution data includes area number, electric vehicle path density, and charging station usage frequency. The hotspot traffic area data includes intersection aggregation density, high-frequency path area number, and hotspot area spatial range. The urban hotspot area number list includes area number with insufficient facility coverage, hotspot area spatial location, and traffic intensity level identifier. The urban charging facility deployment path information includes boundary path intersection points, deployment candidate spatial locations, and connecting path topology. The urban charging hotspot area prediction and planning results include path intersection node numbers, time period matching trajectories, and hotspot area priority sequences.

[0006] As a further aspect of the present invention, the aggregation density value refers to the spatial concentration of the intersection points of different electric vehicle travel paths within a unit area in the urban grid, which measures the density of traffic hotspots.

[0007] As a further aspect of the present invention, the identification of areas with insufficient charging facility coverage refers to locating the area identification code by comparing the distribution of hot traffic areas with existing charging stations.

[0008] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Obtain the urban traffic flow layer, electric vehicle travel path layer, and charging station usage distribution layer, number them according to the urban grid division rules, extract the geometric information of the grid numbers in the three layers, perform number association based on the positional overlap relationship, and generate a layer number matching table. S102: Call the layer number matching table, extract the three types of layer data frames corresponding to the overlapping numbers, perform field concatenation in the grid number dimension, construct the regional index structure, and generate a regional number joint dataset; S103: Based on the numbers and fields in the region numbered joint dataset, aggregate traffic status, traffic trajectory density and charging station usage frequency, perform weighted and normalized processing, construct a spatial matrix, and obtain spatial charging behavior distribution data.

[0009] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Based on the spatial charging behavior distribution data, extract the coordinate sequence of all location points in the electric vehicle path, identify the intersection coordinates between adjacent trajectory segments, statistically count the intersection coordinates within the urban grid, construct the mapping relationship between grid number and the number of intersection points, and obtain the path intersection point aggregation count table. S202: Call the path intersection aggregation count table, calculate the ratio between the number of intersections in the grid number and the unit area, use the ratio as the grid aggregation density value, establish a structured table including grid number, area and density value, and generate a path aggregation density matrix; S203: Based on the density value field in the path aggregation density matrix, a preset density filtering threshold is used as the density differentiation benchmark value. Grid numbers with density values ​​lower than the benchmark value are filtered out, and only the set of numbers corresponding to the density area is retained. The boundary of the area covered by the set of numbers is aggregated to obtain hotspot traffic area data.

[0010] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Based on the hotspot traffic area data, extract the set of grid numbers corresponding to the area boundary, obtain the spatial range of the area covered by all numbers, and filter all charging station location points within the spatial range from the charging station data to perform coordinate clustering and labeling, and establish a charging station location information table in the hotspot area. S302: Call the charging station location information table in the hot spot area, calculate the ratio of the number of charging stations to the area area in each hot spot area, use the ratio as the unit area coverage index, compare it with the preset charging station coverage threshold, identify the set of numbers with coverage index below the threshold, and generate a list of numbers with insufficient charging coverage. S303: Based on the list of insufficient charging coverage numbers, retrieve the area attribute field corresponding to the number, deduplicate all numbers that meet the conditions and aggregate them into a set, establish a list of number fields with a standard structure format, and obtain the list of urban hotspot area numbers.

[0011] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Based on the list of urban hotspot area numbers, retrieve the map boundary vector data corresponding to the number, and extract the coordinates of the intersection path nodes and the set of coordinates of the preset deployment candidate points within each boundary graphic. Perform spatial location aggregation processing on the node coordinates and point coordinates to obtain the boundary path intersection and candidate point table. S402: Call the boundary path intersection and candidate point table, extract the coordinate sequence of connecting line segments between intersection points within the boundary, and combine and collect the line segments according to the number. Analyze the adjacency of endpoints and intermediate points in the line segment set, identify line segment chains that can form closed and continuous paths, and generate a continuous path structure table within the area. S403: Based on the line segment order relationship in the continuous path structure table within the area and the spatial distribution of deployment candidate points, the candidate points on each continuous path are marked with path codes, and a set of path number, coordinate sequence and candidate point index fields is established to obtain the deployment path information of urban charging facilities.

[0012] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: Based on the urban charging facility deployment path information and electric vehicle movement trajectory data for different time periods, extract the coordinate sequence and trajectory point set on the path segment, perform position matching between trajectory points and path coordinates under the same coordinate system, record the spatial index of the overlapping area, and obtain the location set of time period intersection nodes. S502: Call the set of cross-node positions for the time period, and encode the cross-nodes according to the path number and coordinate order fields based on the position of the matching coordinates in the path. Arrange all cross-nodes in the logical order of the path to generate a path node encoding sequence. S503: Based on the cross-density and time period distribution characteristics in the path node coding sequence, statistically analyze the high-frequency coding intervals on the path segment, extract the corresponding spatial segment number and node density value, construct a table structure including path number, cross-weight, and time period coverage information, and obtain the prediction and planning results of urban charging hotspot areas.

[0013] As a further aspect of the present invention, the identification of line segment chains that can form closed and continuous paths refers to filtering a set of line segments that can be connected end-to-end and sequentially to form a path by analyzing the spatial adjacency relationship between the endpoints and intermediate points of the line segments within the boundary.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, a regional behavior model is constructed by integrating traffic flow, electric vehicle routes, and charging usage data. Hotspot areas are identified by combining path intersection density to improve the positioning accuracy of high-demand areas. A coverage identification mechanism is introduced to enhance the targeting of site selection. Deployment paths are extracted based on boundary structure to enhance the continuity and scalability of the layout. Intersection nodes are extracted and encoded by combining time-period trajectories to achieve deployment adaptability to dynamic changes and improve the rationality and practicality of planning results. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] Please see Figure 1 This invention provides a method for predicting and planning urban charging hotspot areas based on big data, including the following steps: S1: By acquiring the urban traffic flow layer, electric vehicle travel path layer, and charging station usage distribution layer, a three-layer fusion table based on area number is established. The overlapping positions of the layers are numbered and integrated according to the urban grid division rules, and spatial charging behavior distribution data is output. S2: Based on spatial charging behavior distribution data, calculate the aggregation density of electric vehicle path intersections in the urban grid, perform spatial filtering based on density values, and generate hotspot traffic area data. S3: Match the location of hotspot traffic area data with the charging station distribution records within the traffic area, identify the area number with insufficient charging facility coverage, and output a list of urban hotspot area numbers. S4: Based on the map boundaries of the areas corresponding to the list of urban hotspot area numbers, extract the number of boundary path intersections and the spatial location of deployment candidates, construct continuous paths based on the geometric relationship of the connecting lines within the boundaries, and generate urban charging facility deployment path information; S5: Match the deployment path information of urban charging facilities with the movement trajectory data of electric vehicles in different time periods, extract the time period intersection nodes and encode them in the order of the path, and output the prediction and planning results of urban charging hotspot areas.

[0023] The three-layer fusion table includes a traffic flow number layer, an electric vehicle route number layer, and a charging station usage number layer. Spatial charging behavior distribution data includes area numbers, electric vehicle route density, and charging station usage frequency. Hotspot traffic area data includes intersection aggregation density, high-frequency route area numbers, and hotspot area spatial range. The urban hotspot area number list includes area numbers with insufficient facility coverage, hotspot area spatial location, and traffic intensity level identifiers. Urban charging facility deployment path information includes boundary path intersections, deployment candidate spatial locations, and connecting path topology. The urban charging hotspot area prediction and planning results include path intersection node numbers, time period matching trajectories, and hotspot area priority sequences.

[0024] Please see Figure 2 The specific steps of S1 are as follows: S101: Obtain the urban traffic flow layer, electric vehicle travel path layer, and charging station usage distribution layer, number them according to the urban grid division rules, extract the geometric information of the grid numbers in the three layers, perform number association based on the positional overlap relationship, and generate a layer number matching table. Based on the need to acquire three types of spatial data—urban traffic flow layer, electric vehicle travel path layer, and charging station usage distribution layer—a unified numbering system is first constructed using an equidistant grid division method, based on the city's geographical boundaries. According to the actual city size, the entire city can be divided into several 500m × 500m square grid areas, each assigned a unique number. Numbering starts from the top left corner of the city and proceeds row by row from left to right. The spatial coordinate information of all data points or graphic objects in the three layers is obtained sequentially, and the numbered area they fall into is determined. For the traffic flow layer, the grid number is obtained through vector lines or trajectory point data, and the traffic flow value, number of vehicles passing through, and average vehicle speed information within that grid are extracted. For example, grid T0102 records 425 vehicles passing through with an average speed of 35km / h, corresponding to the traffic data field content under this number. For the electric vehicle travel path layer, based on the latitude and longitude of the real-time trajectory points of electric vehicles, the grid number each trajectory point falls into is determined sequentially. A complete trajectory is broken down into multiple consecutive numbered segments, each representing a record of an electric vehicle passing through a corresponding area. If an electric vehicle's path crosses grid number R020... 5. For R0206 and R0207, the path has passage records in three numbered areas respectively. For the charging station usage layer, the spatial location of each charging station is matched to the corresponding grid number, and the usage frequency of charging piles in each numbered area is counted. For example, if there are 3 charging stations in the area corresponding to number C0312, and the average daily total number of charging behaviors is 21, then the usage frequency data corresponding to this number is recorded. After extracting the above three types of number information, it is necessary to perform spatial location overlap judgment between grids. For all traffic layers, passage path layers, and charging station layer numbered areas, the judgment is performed in sequence. To determine whether their boundaries overlap, we judge each condition based on whether their spatial positions coincide, the area of ​​the boundary intersection, and the distance between their spatial center points. In the example, if the overlapping area of ​​the boundaries of numbers T0340, R0340, and C0340 exceeds 10% of the total area, and the distance between their spatial center points is less than 300 meters, then we determine that there is a valid spatial overlap relationship between the three numbers, and record it as a number group (T0340, R0340, C0340). We extract all valid number combinations in this way, and finally generate a layer number matching table composed of number triplets for subsequent data linking processing.

[0025] S102: Call the layer number matching table, extract the three types of layer data frames corresponding to the overlapping numbers, perform field concatenation in the grid number dimension, construct the regional index structure, and generate the regional number joint dataset; Read each group of numbers containing three types of layer numbers row by row. Perform data extraction and field join operations on each group of numbers. Taking the number group (T0340, R0340, C0340) as an example, retrieve the traffic data field corresponding to number T0340 from the traffic flow layer. For example, the number of vehicles passing through is 480, the average speed is 30km / h, and the traffic density is 8 vehicles per minute. Extract the trajectory information corresponding to number R0340 from the electric vehicle travel path layer, including the number of travel paths recorded in the area, the average path length, and the path point density. If there are 97 travel paths recorded in the area and the average spacing between path points is 20 meters, the path density under this number can be estimated by dividing the number of path points by the area. At the same time, extract the charging pile information corresponding to number C0340 from the charging station usage layer. For example, if there are 5 charging piles, the average number of charging times per day is 40, and the peak time is concentrated at 16. Between 00:00 and 20:00, after the above data is extracted, the three types of data are joined under the grid number dimension according to the corresponding relationship in the layer number matching table to construct a complete area number data structure. The original number identifier and the extracted field value are retained in the field, and the layer identifier prefix is ​​added to the field with the same name. For example, the traffic flow field is named T_flow, the path density field is named R_density, and the charging frequency field is named C_freq. After joining all number groups, each number combination and its joining fields are organized into a structured record. An index structure is created according to the area number to improve the efficiency of subsequent queries. Each joint number can be concatenated using a unified encoding rule. For example, T0340_R0340_C0340 is abbreviated to area number Z0340. This area number is used as the key value to correspond to the complete field data, and finally a complete area number joint dataset is formed.

[0026] S103: Based on the numbers and fields in the region number joint dataset, aggregate traffic status, traffic trajectory density and charging station usage frequency, perform weighting and normalization processing, construct a spatial matrix, and obtain spatial charging behavior distribution data; The values ​​of the fields corresponding to each region number are read one by one. For the three fields T_flow, R_density, and C_freq, numerical classification and aggregation are performed respectively. For the traffic status field T_flow, after extracting the traffic flow value, its value is divided into three levels according to a preset range. If the number of vehicles passing through a certain region is 480, it falls into the medium flow level range, and its corresponding level value is set to 0.5. For the path density field R_density, the number of paths or the total number of path points per unit area is calculated. The density value is obtained by dividing the total number of passing path points by the grid area. If the total number of path points in a certain region is 2100 and the grid area is 0.25 square kilometers, the path density is 8400 points / km², falling into the dense level, and the corresponding aggregation value is set to 0.9. For the C_freq field, based on each region number... The daily usage frequency of charging stations within the area is categorized. If a certain area has an average daily usage frequency of 24 times, which is considered a medium-frequency level, the aggregation value is set to 0.55. After obtaining the aggregation level values ​​of the three indicators, they are weighted according to the weight configuration. For example, the weight of traffic status is 0.4, the weight of path density is 0.35, and the weight of charging frequency is 0.25. The comprehensive aggregation value is calculated for each numbered area. The obtained value is then normalized by the maximum and minimum values ​​under all area numbers. Assuming that the maximum comprehensive value is 0.91 and the minimum is 0.31 among all numbers, the aggregation value of a certain numbered area is 0.6525, and the corresponding normalized value is 0.571. Finally, the normalized values ​​of all numbered areas are mapped to the spatial grid of the corresponding location to form a matrix-like spatial structure. Each grid cell records its normalized comprehensive value of charging behavior, which is then summarized to form a spatial charging behavior distribution data matrix.

[0027] Please see Figure 3 The specific steps of S2 are as follows: S201: Based on spatial charging behavior distribution data, extract the coordinate sequence of all location points in the electric vehicle path, identify the intersection coordinates between adjacent trajectory segments, statistically count the intersection coordinates within the urban grid, construct the mapping relationship between grid number and the number of intersection points, and obtain the path intersection point aggregation count table. The collected electric vehicle travel paths are processed one by one, breaking each path down into a sequence of location coordinates arranged chronologically. These coordinates consist of both longitude and latitude values, forming a continuous set of points in the recording order. Adjacent location points are then connected pairwise to form trajectory segments. These trajectory segments within each path are then compared and combined sequentially. For adjacent trajectory segments within the same path, the coordinates of their starting and ending points are read to determine if the two trajectory segments intersect in planar projection. When two trajectory segments overlap simultaneously in both the horizontal and vertical coordinate ranges and are in opposite directions, an intersection is identified, and this intersection is recorded as a valid intersection point. For example, a path might contain two trajectory segments: one from point A to point B, and the other from point C to point D. If two segments intersect within the same coordinate range, the intersection point is recorded as one intersection. After identifying the intersection points of a single path, the above process is repeated for all paths to form a complete set of intersection coordinates. Then, each intersection coordinate is matched with a city grid. By determining whether the latitude and longitude of the intersection coordinate falls within the spatial boundary corresponding to a certain grid number, its grid number is determined. Multiple intersection coordinates appearing in the same grid are counted and accumulated. For example, if 5 intersection coordinates are identified in grid number G215, the number of intersection points for that number is recorded as 5. After completing the statistics for all grids in the entire city, a mapping relationship is constructed with grid number as index and intersection point count as value. The mapping results are then summarized to form a path intersection point aggregation count table.

[0028] S202: Call the path intersection aggregation count table, calculate the ratio between the number of intersections in the grid number and the unit area, use the ratio as the grid aggregation density value, establish a structured table including grid number, area and density value, and generate the path aggregation density matrix; The system reads the number of intersections corresponding to each grid number one by one, and simultaneously obtains the actual area information of the grid. The grid area is kept consistent with the previously unified division rules. For each number, the ratio of its number of intersections to its corresponding area is calculated to obtain the path intersection aggregation density value within that number area. During the execution process, the number of intersections under the number is read first, and then the area value corresponding to that number is read. Subsequently, the ratio between the number and the area is calculated, and the calculation result is used as the density field value of that number. For example, if a number area has 10 intersections and the area of ​​the area remains fixed, a clear density value is obtained through the correspondence between the number and the area. The three pieces of information, number, area, and density value, are organized into a structured record. The above operation is repeated for all numbers to form a complete structured data set. When organizing the data, the density value is mapped to a two-dimensional spatial structure according to the number order. Each grid number corresponds to a fixed position in the matrix, and the value in the matrix is ​​the path aggregation density value of that number area. The row and column order of the matrix is ​​consistent with the arrangement of the city grid, thereby generating a path aggregation density matrix covering the entire city area.

[0029] S203: Based on the density value field in the path aggregation density matrix, the preset density filtering threshold is used as the density differentiation benchmark value. Grid numbers with density values ​​lower than the benchmark value are filtered out, and only the set of numbers corresponding to the density area is retained. The boundary of the area covered by the set of numbers is aggregated to obtain the hot spot traffic area data. Each grid number is individually evaluated. A pre-set density screening threshold is used as the distinguishing criterion. This threshold is set based on the density distribution of the entire city. For example, if statistics show that most grid densities are concentrated within a certain range, the median of that range is selected as the screening benchmark. During the evaluation process, the density value corresponding to each grid number is read sequentially and compared with the set threshold. If the density value is lower than the threshold, the number is removed from the candidate set. If the density value is not lower than the threshold, the number is retained in the number set. After evaluating all grids, a number set consisting only of high-density grid numbers is obtained. Then, boundary aggregation processing is performed on the spatial area corresponding to this number set. By reading the spatial boundary coordinates of each number, the continuity of adjacent or boundary-touching numbers is evaluated. Numbers that meet the adjacent conditions are merged into the same area. During the merging process, the area boundary is continuously expanded until no new adjacent numbers can be introduced. Finally, a complete external boundary outline is generated for each group of consecutive numbers, and all grid numbers covered within it are recorded, forming multiple area boundary data aggregated from high-density numbers. Finally, a hotspot traffic area data set is obtained.

[0030] Please see Figure 4 The specific steps of S3 are as follows: S301: Based on the data of hot traffic areas, extract the set of grid numbers corresponding to the boundaries of the areas, obtain the spatial range of the area covered by all numbers, and select all charging station location points within the spatial range from the charging station data for coordinate clustering and labeling, and establish a charging station location information table in the hot traffic area. First, the spatial boundary information of each area is analyzed. By reading the set of numbers corresponding to that area, the spatial range covered by all grid numbers under that set is obtained. The boundary coordinate set corresponding to each number is read one by one, and the boundaries of all numbered areas are merged to form the overall boundary outline of the hotspot area. Next, the spatial location coordinates of all charging stations are extracted from the city-wide charging station data. It is then determined whether the coordinates of each charging station fall within the boundary range of the hotspot area. By determining whether the latitude and longitude of the charging station are within the boundary outline of the area, all charging stations that meet the conditions are selected to form a set of charging stations within the hotspot area. After obtaining all the point sets, the coordinates of the charging station locations are clustered. Charging stations that are close to each other are grouped and labeled as the same cluster unit. The spatial straight-line distance between the points is then determined. If the distance between charging stations is less than a set clustering interval threshold (e.g., 200 meters), all charging stations within that distance are grouped into the same category. A unique clustering number is assigned to each cluster. After completing all clustering processes, a clustering label field is added to each charging station in the hotspot area. Subsequently, a complete table of charging station location information within the hotspot area is established. Each record in the table corresponds to a charging station, and the fields include: charging station number, latitude and longitude coordinates, area number, cluster number, and whether it is a boundary station. For example, a station with the number CS-024, coordinates (118.7612, 32.0489), area number HZ102, and cluster number CL08 constitutes a valid record. Records for all stations are processed sequentially to finally form a structured table of charging station location information within the hotspot area.

[0031] S302: Call the charging station location information table in the hot spot area, calculate the ratio of the number of charging stations to the area area in each hot spot area, use the ratio as the coverage index per unit area, compare it with the preset charging station coverage threshold, identify the set of numbers whose coverage index is lower than the threshold, and generate a list of numbers with insufficient charging coverage. The number of charging stations under each zone number is counted. All charging station records corresponding to each zone number are read, and the total number of records grouped by zone number is used as the number of charging stations in that zone. Next, the total area covered by all grid numbers under that zone number is obtained. This area value is obtained by multiplying the number of zones by the area of ​​a single grid. For example, zone number HZ205 consists of 5 grid numbers, and the area of ​​a single grid is 0.25 square kilometers, so the area of ​​that zone is 1.25 square kilometers. If there are 3 charging stations within it, then the coverage index per unit area of ​​that zone is 3 ÷ 1.25 = 2.4 stations / km². This coverage index calculation is performed for all zones. After marking, a unified comparison operation is performed, reading the preset charging station coverage threshold. The threshold is determined by the overall charging density level of the city. For example, if the city-level coverage benchmark is set to no less than 4 charging stations per square kilometer, then 4 is the threshold used for judgment. The coverage index of each area is compared with 4. If the index is less than 4, the area is judged as an area with insufficient coverage, and the area number is recorded into the insufficient coverage number set. For example, the coverage index of number HZ205 is 2.4, which is less than the benchmark value of 4, so it is added to the set. This judgment is performed on all areas in turn, and finally the list of area numbers with coverage index below the threshold is output, forming the charging coverage insufficient number list.

[0032] S303: Based on the list of insufficient charging coverage numbers, retrieve the area attribute field corresponding to the number, deduplicate all numbers that meet the conditions and aggregate them into a set, establish a list of number fields in a standard structure format, and obtain the list of numbers for urban hotspot areas. For each number in the list, its attribute fields in the original city area numbering structure are retrieved one by one. The attribute fields contain information such as the administrative division, function type, street, and transportation type corresponding to the area number. When reading the attribute fields, the record table content is queried for each number in turn. The corresponding detailed attributes are obtained by matching the number index. It is determined whether each number meets the retention conditions. For example, only numbers with the function type of "residential area" or "commercial area" in the attribute field are filtered. The numbers that meet the conditions are marked. All numbers that meet the conditions are deduplicated to remove duplicate numbers and retain only the unique identifier number. After the aggregation process of the number set is completed, all numbers are summarized into a number field set and organized into a number list according to a unified field format. The list field structure includes number, area name, function type, street, aggregation number identifier, etc. Each record in the list corresponds to a number unit. For example, number HZ205, function type is "commercial area", street is "Nanjing Road Street", which constitutes a number field record. Finally, the integrated standard structure format number field list is output as the urban hotspot area number list.

[0033] Please see Figure 5 The specific steps of S4 are as follows: S401: Based on the list of urban hotspot area numbers, retrieve the map boundary vector data of the corresponding number, and extract the coordinates of the intersection path nodes and the set of coordinates of the pre-deployed candidate points within each boundary graphic. Perform spatial location aggregation processing on the node coordinates and point coordinates to obtain the boundary path intersection and candidate point table. The boundary information of each area number is read sequentially. Map boundary vector data corresponding to each number is retrieved from the geographic information system. This vector data contains a set of boundary line segments composed of multiple coordinate points. After extracting the boundary data for each area number, a polygonal graphic object of the area is constructed. Then, the traffic path intersection node information within the boundary graphic is read. The coordinates of the intersection nodes are obtained from the previously identified path intersection point coordinate table. When acquiring node data, each node's coordinates are checked to see if they are within the area boundary polygon. The node's affiliation is confirmed through spatial location determination. Simultaneously, all points within the boundary area are extracted from a pre-set set of candidate deployment points. The candidate deployment point coordinates consist of urban traffic corridors, existing power distribution capacity sites, and vacant land. If the latitude and longitude of each point fall within the current area boundary, it is retained in the candidate point set of this area. Then, the extracted intersection path nodes and candidate deployment points are processed for spatial location aggregation. Each intersection node is traversed, and with the point as the center, it is searched for whether there is a candidate deployment point within its surrounding radius r (e.g., 200 meters). If there is, the candidate point and the intersection node are paired to form a spatial relationship. If no candidate point falls within the range, the intersection node is not recorded as a pair. All nodes are traversed in this way, and the information of all node pairs with intersection relationships is recorded. Each pair of records includes fields such as intersection path node number, coordinates, matched candidate deployment point number, deployment point coordinates, and distance value. All information is organized into a boundary path intersection and candidate point table.

[0034] S402: Call the boundary path intersection and candidate point table, extract the coordinate sequence of connecting line segments between intersection points within the boundary, and combine and collect the line segments according to the number. Analyze the adjacency of endpoints and intermediate points in the line segment set, identify line segment chains that can form closed and continuous paths, and generate a continuous path structure table within the area. For each hotspot area, connecting line segments are extracted between intersections. The coordinates of all identified intersections are read. For each pair of intersections with a path connection, the resulting path segment is extracted. The path segment is represented by a continuous sequence of trajectory coordinates between two points. If intersections A and B exist within area HZ103, and the path structure contains an actual travel trajectory connecting A and B, then the path point sequence from A to B is extracted as a line segment. The starting point coordinates, ending point coordinates, and all intermediate interpolation points are recorded. After extracting the line segments, all path segments within that area are grouped and aggregated using the area number as the primary key, forming a line segment set table for that area. Subsequently, a topological analysis is performed on each line segment to determine the connection between the endpoints and intermediate points of each line segment. The adjacency relationship is determined by sequentially reading the start and end points of each line segment and counting whether the endpoint is the start or end point of other line segments. If a point is adjacent to two or more line segments, it is recorded as an intersection node in the network topology. If a point is connected to only one line segment, it is a boundary endpoint. Then, all line segments are chained together according to the connection relationship between them. If several line segments can be connected sequentially from the start point to the end point, and the start and end points coincide, a closed path is formed. If the start and end points do not coincide, but the intermediate points are continuously connected in the preceding and following paths, a continuous path is formed. All closed paths and continuous path structures are recorded in order to form a sequence table of line segment numbers and an endpoint list for each path, and finally, a continuous path structure table within the area is formed.

[0035] S403: Based on the line segment order relationship in the continuous path structure table within the area and the spatial distribution of deployment candidate points, the candidate points on each continuous path are marked with path codes, and a set of path number, coordinate sequence and candidate point index fields are established to obtain the deployment path information of urban charging facilities. The process involves sequentially reading the line segment numbers and endpoint coordinates of each continuous path, treating this path as a complete travel route for deployment point matching, reading the coordinate range of all line segments on the path, establishing a continuous coordinate sequence to represent the path trajectory, filtering the set of deployment candidate points, identifying candidate deployment points whose adjacent distances within the path coordinate range are within a set radius, and determining that a candidate point belongs to the path's deployment candidate point if its coordinates are less than 100 meters from the path trajectory line. Each qualified candidate point is then marked with a path code and designated as a deployment node under the current path number. During the marking process, a path number and its corresponding coordinates are constructed. The index relationship between candidate point numbers is recorded, and the position index information of the point in the path coordinate sequence is recorded. For example, if the path number is P_HZ103_05 and the corresponding candidate point number is D017, then the sequence position of D017 in the path P_HZ103_05 is recorded as the end of the 4th segment. All candidate points are uniformly organized according to their path belonging relationship and the position sequence information within the path to form a set of structured fields. The fields include path number, complete coordinate sequence, deployment point number, point coordinates, and the position index of the point in the path, etc., and finally form the deployment path information of charging facilities in multiple hot spots within the city.

[0036] Please see Figure 6 The specific steps of S5 are as follows: S501: Based on the urban charging facility deployment path information and electric vehicle movement trajectory data at different time periods, extract the coordinate sequence and trajectory point set on the path segment, perform position matching between trajectory points and path coordinates under the same coordinate system, record the spatial index of overlapping areas, and obtain the location set of time period intersection nodes. First, the system reads each path number and its corresponding path segment coordinate sequence recorded in the deployment path information. Each path segment is expanded and processed, converting its trajectory coordinate sequence into a continuous set of spatial line segment coordinates. Then, the system sequentially reads the electric vehicle movement trajectory point data for different time periods, classifying the trajectory points according to their respective time period labels (e.g., morning peak 07:00–09:00, off-peak 10:00–15:00, evening peak 17:00–19:00) to ensure that each type of trajectory data corresponds to a specific time period. Next, the system extracts the latitude and longitude coordinates of each trajectory point and converts them into XY position points in a unified coordinate system. After completing the standardized spatial positioning of all trajectory points, a spatial location matching operation is performed on each coordinate point in the path segment coordinate sequence. During the matching process… For each path segment's coordinates, a spatial buffer radius is set, for example, 20 meters. All electric vehicle trajectory points within this buffer range are searched, and the correspondence between these trajectory points and path segment points is recorded. If there are 5 trajectory points within 20 meters of the 18th point of a path segment P_HZ120, it is considered a spatial intersection. This operation is repeated for all path segments, accumulating and recording all spatially overlapping locations. These overlapping points are categorized and stored according to path numbers. Each path corresponds to a time period intersection node set. Each node in the set records the time period to which the trajectory point belongs, the path segment coordinate point number, the spatial location index, the number of matching trajectory points within the buffer radius, and other field information. Finally, the time period intersection node location set generated after matching all path segments with multi-time period trajectory data within the entire city is obtained.

[0037] S502: Call the time period cross node location set, and encode the cross nodes according to the path number and coordinate order fields based on the position of the matching coordinates in the path. Arrange all cross nodes according to the logical order of the path to generate a path node encoding sequence. For each path number, the intersection nodes are further encoded and organized. First, the intersection nodes are grouped according to the path number to ensure that each group of intersection nodes corresponds to one path number. Then, within each group, the original position order of each intersection node in the path segment coordinate sequence is read. By comparing the spatial index position of the intersection node with the path coordinate point number, its relative position in the path segment is determined, and this position is used as the sorting criterion for ascending order. If the coordinate sequence of path P_HZ120 has 50 points, and its intersection nodes fall on the 6th, 12th, 14th, 21st, and 39th coordinate points, then the sorted order should be 6→12→14→21→39. The sorting results are prefixed with the path number, and a unique identifier is generated for each intersection node according to the node order. For example, the node number can be set as P_HZ120_N06, P_HZ120_N12, P_HZ120_N14, etc. The path ownership information, sequence number in the path, coordinate position value, corresponding time period label, etc. of each intersection node are summarized to form a record table. The record table is organized with the path number as the primary key structure, and each record is an intersection node entry in the path. Finally, a path node coding sequence is formed. Each path corresponds to a complete node coding list. The node coding can be used to describe the temporal mapping relationship between the actual location of the electric vehicle on the path and the deployed path structure.

[0038] S503: Based on the cross density and time period distribution characteristics in the path node coding sequence, statistically analyze the high-frequency coding intervals on the path segment, extract the corresponding spatial segment number and node density value, construct a table structure including path number, cross weight, and time period coverage information, and obtain the prediction and planning results of urban charging hotspot areas. The density and time-period distribution of nodes in the path segment are identified. First, each path is traversed and counted. The path coordinate sequence is divided into windows, with each window consisting of 10 coordinate points. The number of intersecting nodes in each window is counted. If there are a total of 9 intersecting nodes recorded in windows 3 to 12, the node density value of this segment is calculated to be 0.9. Then, the node density of each segment is compared with a set density interval. For example, a high-density interval threshold is set as a segment with a node density of not less than 0.8. If the segment meets the condition, it is marked as a high-frequency segment. The path number and start and end coordinate indices corresponding to this segment are recorded as identifiers. Then, the time period index of the intersecting nodes in the high-frequency segment is counted. The distribution of nodes within each segment is recorded, including the number of time periods and their distribution percentage. For example, if 4 out of 9 nodes in a segment are during the morning peak, 3 during the off-peak, and 2 during the evening peak, the time period coverage information for that segment is recorded as follows: morning peak 44.4%, off-peak 33.3%, and evening peak 22.2%. Table records are created for each segment that meets the high-frequency standard. The field structure includes path number, segment number, cross weight (i.e., the number of cross nodes per unit length), node density value, time period label, and corresponding percentage. By statistically summarizing all path segments, a table data structure for predicting and planning charging hotspot areas across the entire city is constructed, ultimately obtaining the prediction and planning results for urban charging hotspot areas.

[0039] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of protection of the technical solution.

Claims

1. A method for predicting and planning urban charging hotspot areas based on big data, characterized in that, Includes the following steps: S1: Obtain the urban traffic flow layer, electric vehicle travel path layer and charging station usage distribution layer, construct a fusion table of the three types of layers based on the area number, and perform numbering integration on the overlapping positions of the layers according to the urban grid division rules to generate spatial charging behavior distribution data; S2: Based on the spatial charging behavior distribution data, calculate the aggregation density of electric vehicle path intersections in the urban grid, and generate hotspot traffic area data by filtering according to the aggregation density value; S3: Match the location of the hotspot traffic area data with the charging station distribution records within the traffic area, identify the area number with insufficient charging facility coverage, and output a list of urban hotspot area numbers. S4: Based on the map boundaries of the corresponding areas in the list of urban hotspot area numbers, extract the number of boundary path intersections and the spatial location of deployment candidates, construct continuous paths based on the geometric relationship of the connecting lines within the boundaries, and generate urban charging facility deployment path information; S5: Match the location of the urban charging facility deployment path information with the electric vehicle movement trajectory data of different time periods, extract the time period intersection nodes and encode them according to the path order, and output the prediction and planning results of urban charging hotspot areas.

2. The method for predicting and planning urban charging hotspot areas based on big data according to claim 1, characterized in that, The three-layer fusion table includes a traffic flow number layer, an electric vehicle path number layer, and a charging station usage number layer. The spatial charging behavior distribution data includes area number, electric vehicle path density, and charging station usage frequency. The hotspot traffic area data includes intersection aggregation density, high-frequency path area number, and hotspot area spatial range. The urban hotspot area number list includes area number with insufficient facility coverage, hotspot area spatial location, and traffic intensity level identifier. The urban charging facility deployment path information includes boundary path intersection points, deployment candidate spatial locations, and connecting path topology. The urban charging hotspot area prediction and planning results include path intersection node numbers, time period matching trajectories, and hotspot area priority sequence.

3. The method for predicting and planning urban charging hotspot areas based on big data according to claim 1, characterized in that, The aggregation density value refers to the spatial concentration of intersections of different electric vehicle travel paths within a unit area in an urban grid, measuring the density of traffic hotspots.

4. The method for predicting and planning urban charging hotspot areas based on big data according to claim 1, characterized in that, The area code for identifying areas with insufficient charging infrastructure refers to the identification code for areas with dense traffic but scarce charging facilities, which is determined by comparing the distribution of existing charging stations with hot traffic areas.

5. The method for predicting and planning urban charging hotspot areas based on big data according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain the urban traffic flow layer, electric vehicle travel path layer, and charging station usage distribution layer, number them according to the urban grid division rules, extract the geometric information of the grid numbers in the three layers, perform number association based on the positional overlap relationship, and generate a layer number matching table. S102: Call the layer number matching table, extract the three types of layer data frames corresponding to the overlapping numbers, perform field concatenation in the grid number dimension, construct the regional index structure, and generate a regional number joint dataset; S103: Based on the numbers and fields in the region numbered joint dataset, aggregate traffic status, traffic trajectory density and charging station usage frequency, perform weighted and normalized processing, construct a spatial matrix, and obtain spatial charging behavior distribution data.

6. The method for predicting and planning urban charging hotspot areas based on big data according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Based on the spatial charging behavior distribution data, extract the coordinate sequence of all location points in the electric vehicle path, identify the intersection coordinates between adjacent trajectory segments, statistically count the intersection coordinates within the urban grid, construct the mapping relationship between grid number and the number of intersection points, and obtain the path intersection point aggregation count table. S202: Call the path intersection aggregation count table, calculate the ratio between the number of intersections in the grid number and the unit area, use the ratio as the grid aggregation density value, establish a structured table including grid number, area and density value, and generate a path aggregation density matrix; S203: Based on the density value field in the path aggregation density matrix, a preset density filtering threshold is used as the density differentiation benchmark value. Grid numbers with density values ​​lower than the benchmark value are filtered out, and only the set of numbers corresponding to the density area is retained. The boundary of the area covered by the set of numbers is aggregated to obtain hotspot traffic area data.

7. The method for predicting and planning urban charging hotspot areas based on big data according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Based on the hotspot traffic area data, extract the set of grid numbers corresponding to the area boundary, obtain the spatial range of the area covered by all numbers, and filter all charging station location points within the spatial range from the charging station data to perform coordinate clustering and labeling, and establish a charging station location information table in the hotspot area. S302: Call the charging station location information table in the hot spot area, calculate the ratio of the number of charging stations to the area area in each hot spot area, use the ratio as the unit area coverage index, compare it with the preset charging station coverage threshold, identify the set of numbers with coverage index below the threshold, and generate a list of numbers with insufficient charging coverage. S303: Based on the list of insufficient charging coverage numbers, retrieve the area attribute field corresponding to the number, deduplicate all numbers that meet the conditions and aggregate them into a set, establish a list of number fields with a standard structure format, and obtain the list of urban hotspot area numbers.

8. The method for predicting and planning urban charging hotspot areas based on big data according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Based on the list of urban hotspot area numbers, retrieve the map boundary vector data corresponding to the number, and extract the coordinates of the intersection path nodes and the set of coordinates of the preset deployment candidate points within each boundary graphic. Perform spatial location aggregation processing on the node coordinates and point coordinates to obtain the boundary path intersection and candidate point table. S402: Call the boundary path intersection and candidate point table, extract the coordinate sequence of connecting line segments between intersection points within the boundary, and combine and collect the line segments according to the number. Analyze the adjacency of endpoints and intermediate points in the line segment set, identify line segment chains that can form closed and continuous paths, and generate a continuous path structure table within the area. S403: Based on the line segment order relationship in the continuous path structure table within the area and the spatial distribution of deployment candidate points, the candidate points on each continuous path are marked with path codes, and a set of path number, coordinate sequence and candidate point index fields is established to obtain the deployment path information of urban charging facilities.

9. The method for predicting and planning urban charging hotspot areas based on big data according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Based on the urban charging facility deployment path information and electric vehicle movement trajectory data for different time periods, extract the coordinate sequence and trajectory point set on the path segment, perform position matching between trajectory points and path coordinates under the same coordinate system, record the spatial index of the overlapping area, and obtain the location set of time period intersection nodes. S502: Call the set of cross-node positions for the time period, and encode the cross-nodes according to the path number and coordinate order fields based on the position of the matching coordinates in the path. Arrange all cross-nodes in the logical order of the path to generate a path node encoding sequence. S503: Based on the cross-density and time period distribution characteristics in the path node coding sequence, statistically analyze the high-frequency coding intervals on the path segment, extract the corresponding spatial segment number and node density value, construct a table structure including path number, cross-weight, and time period coverage information, and obtain the prediction and planning results of urban charging hotspot areas.

10. The method for predicting and planning urban charging hotspot areas based on big data according to claim 8, characterized in that, The identification of line segment chains that can form closed and continuous paths refers to filtering the set of line segments that can be connected end-to-end and sequentially to form a path by analyzing the spatial adjacency relationship between the endpoints and intermediate points of the line segments within the boundary.

Images