Charging pile intelligent site selection evaluation method and system

Abstract

The invention provides an intelligent charging pile site selection evaluation method and system. The method comprises the following steps: acquiring communication data of a mobile terminal in a target area in real time, and constructing a demand time period division model to distinguish daytime quick charging and night conventional charging demands; dynamically adjusting the position of an edge computing device based on demand distribution to ensure that the service range is matched with a people flow density peak region in real time; controlling the multi-port wireless charging system to switch an adaptive charging power mode according to the identified charging demand type; and generating a candidate charging pile layout position set in combination with equipment layout and power mode data, and screening out an optimal deployment scheme through safety distance verification. According to the technical scheme provided by the invention, efficient matching between the charging pile layout and the dynamic people flow demand is realized, and the daytime charging efficiency and the night equipment utilization rate are improved.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent site selection and evaluation, and in particular to an intelligent site selection and evaluation method and system for charging piles. Background Art

[0002] There is a significant difference in charging needs between daytime office use and nighttime residential use in mixed-function urban areas (such as commercial and residential complexes and properties above subway stations). During the day, the short-term high-power commercial vehicle rapid charging needs must be met, while at night, the long-term regular charging needs of residential vehicles must be met. Such scenarios require that the layout of charging piles be able to respond to dynamic changes in passenger flow in real time and achieve efficient allocation of charging resources within a limited space. At the same time, it is necessary to avoid conflicts in underground pipe networks and balance grid load fluctuations. This poses a comprehensive technical challenge to dynamic demand perception, coordinated optimization of spatial resources, and energy efficiency adaptation.

[0003] The current mainstream solution is based on a static charging demand prediction model. By integrating historical charging order data with fixed time division rules (such as working days / holidays), combined with the grid capacity distribution data of the geographic information system (GIS), a multi-objective optimization algorithm is used to generate the charging pile layout location. This solution sets a fixed power allocation ratio between daytime and nighttime, and uses a genetic algorithm or a particle swarm optimization algorithm to search for a site selection solution that meets the grid capacity constraints and maximizes coverage within a preset geographical range.

[0004] The existing solutions have the following defects: the static model cannot capture real-time fluctuations in passenger density (such as a surge in charging demand caused by sudden commercial activities), resulting in coexistence of equipment overload during the day and idle resources at night; the fixed power allocation mode ignores the differences in power characteristics of charging demand at different time periods, making it difficult to achieve high-efficiency operation (such as conflicts between daytime fast charging demand and peak load of the power grid); relying on offline optimization results, it is impossible to dynamically adjust the equipment position to avoid newly discovered underground pipeline conflicts during construction, and frequent manual intervention is required to correct the solution. Summary of the invention

[0005] The present application provides a charging pile intelligent site selection and evaluation method and system to solve the problems of low daytime charging efficiency and low nighttime equipment utilization in the prior art.

[0006] In a first aspect, the present application provides a charging pile intelligent site selection and evaluation method, comprising: Acquire mobile terminal communication data of the target area, and generate a demand period division model based on the spatiotemporal variation law of the crowd density in the mobile terminal communication data, wherein the demand period division model is used to distinguish between the fast charging demand corresponding to the daytime period and the conventional charging demand corresponding to the nighttime period; According to the demand distribution result output by the demand period division model, the position of the edge computing device deployed in the target area is adjusted so that the service range of the edge computing device moves in a consistent manner with the peak area of ​​the pedestrian flow density; Based on the charging requirement type identified in the position adjustment result, controlling the multi-port wireless charging system to switch the charging power output mode, the charging power output mode corresponding to the time period of the fast charging requirement and the regular charging requirement; Combining the network layout data after the position adjustment of the edge computing device with the information corresponding to the power output mode of the multi-port wireless charging system, generating a set of candidate charging pile layout locations; A distance verification is performed on each position in the candidate charging pile layout position set, and a target charging pile layout scheme that meets the preset safety standard and matches the demand distribution law of the demand period division model is screened out.

[0007] Optionally, the combining the network layout data after the position adjustment of the edge computing device and the power output mode information of the multi-port wireless charging system to generate a set of candidate charging pile layout locations includes: extracting the network layout data after the position adjustment of the edge computing device, and identifying a continuous spatial area that meets the daytime charging service efficiency requirements; synchronously acquiring power output mode information of the multi-port wireless charging system, and marking output areas where the power utilization rate during the night time period is continuously higher than the average level; Performing geospatial superposition of the continuous spatial region and the output region, and screening out candidate layout ranges that include both the daytime efficiency region boundary and the nighttime utilization rate core region; Based on the candidate deployment range, the deployable areas are spatially prioritized according to a preset charging pile spacing rule to generate a set of candidate charging pile deployment locations.

[0008] Optionally, the performing distance verification on each position in the candidate charging pile layout position set to screen out a target charging pile layout scheme that meets a preset safety standard and matches a demand distribution law of the demand period division model includes: Obtaining the spatial distance between each location in the candidate charging pile layout location set and the three-dimensional coordinate data of the underground pipe network, and screening out a subset of each location whose straight-line distance to the gas pipeline and the power pipeline gallery is greater than a preset safety distance; Based on the spatial distribution of the fast charging demand hotspots and the conventional charging demand hotspots marked in the demand period division model, the area ratio of the subset covering the two types of demand hotspots is calculated, and the adaptation position that retains the area ratio while meeting the lower limit of the coverage of the daytime fast demand and the nighttime conventional demand is retained; Verify the underground pipe network safety spacing of the adaptation position, eliminate abnormal positions whose spacing with adjacent pipe network positions exceeds the preset spacing interval, and generate a set of intermediate layout plans; A safety assessment is performed based on the historical time period passenger density curve in the intermediate layout plan set, and the test results are similarly matched with the demand distribution law of the demand period division model to screen out the target charging pile layout plan that meets the preset safety standards and matches the demand distribution law.

[0009] Optionally, adjusting the location of the edge computing device deployed in the target area according to the demand distribution result output by the demand time period division model includes: Analyze the demand distribution result output by the demand period division model, extract the high demand area coordinate set in the target area in the current period, and determine the moving trajectory of the peak area of ​​the pedestrian flow density according to the spatiotemporal distribution characteristics of the high demand area coordinate set; Based on the center position of the peak area of ​​each time segment in the moving trajectory, a target deployment position set of the edge computing device is calculated, wherein the straight-line distance between each target position in the target deployment position set and the center of the peak area of ​​the corresponding time segment does not exceed a preset coverage radius; Generate a device migration instruction according to the target deployment location set, drive the edge computing device to migrate to the target location according to the device migration instruction, and establish a communication network with adjacent devices during the migration process, and control the end-to-end transmission delay between nodes in the communication network to be lower than a preset maximum value; An updated coordinate set of the peak area is obtained through the communication network. When it is detected that the coverage overlap rate between the updated coordinate set and the target deployment position set is lower than a preset tolerance threshold, the target deployment position set calculation is re-executed based on the updated coordinate set, and the device migration instruction is synchronously adjusted.

[0010] Optionally, the controlling the multi-port wireless charging system to switch the charging power output mode based on the charging requirement type identified in the position adjustment result includes: Acquire a charging requirement type marked in the position adjustment result, where the charging requirement type includes a first type of demand signal and a second type of demand signal corresponding to the fast charging requirement and the conventional charging requirement, respectively; According to the spatial distribution density of the charging demand type, a priority matching rule is configured for each charging port of the multi-port wireless charging system, and the priority of the charging port in the area covered by the first type of demand signal is higher than that of the second type of demand signal; Based on the priority matching rule, adjusting the electromagnetic coupling parameter combination of each charging port in the multi-port wireless charging system so that the output power range of the high-priority charging port is adapted to the fast charging requirement, and the output power range of the low-priority charging port is adapted to the conventional charging requirement, wherein the electromagnetic coupling parameter combination includes a resonant frequency matching range and a power transmission coil spacing; The electromagnetic coupling state feedback data of the multi-port wireless charging system is used to periodically detect the degree of compatibility between the actual output power of each charging port and the priority matching rule. If it is detected that the degree of compatibility is lower than the matching rule requirement, the electromagnetic coupling parameter combination is regenerated and the priority matching rule is updated.

[0011] Optionally, the generating a demand time period division model based on the spatiotemporal variation law of the crowd flow density in the mobile terminal communication data includes: Extracting the spatial coordinates and timestamps of the terminal devices in the mobile terminal communication data, and constructing a spatiotemporal distribution sequence of the density of human traffic with geographic grids as units; Performing pattern recognition on the fluctuation characteristics of the crowd density in each geographic grid in the crowd density spatiotemporal distribution sequence, extracting a set of grids with short-term high-density aggregation characteristics during the daytime period as candidate areas for fast charging demand, and extracting a set of grids with continuous medium-low density distribution during the nighttime period as candidate areas for conventional charging demand; Based on the spatiotemporal overlap analysis between the candidate areas with fast charging demand and the candidate areas with conventional charging demand, the candidate areas whose spatial coverage areas in adjacent time segments exceed the upper limit of the overlap are merged into demand period division units, and the corresponding charging demand type label is marked for each division unit; According to the label distribution law of the demand time division unit in the historical period, the division granularity of the geographic grid and the determination threshold of the spatiotemporal overlap analysis are modified to generate a demand time division model adapted to the changing characteristics of the flow of people in the target area.

[0012] Optionally, generating a device migration instruction according to the target deployment location set, driving the edge computing device to migrate to the target location according to the device migration instruction, and establishing a communication network with adjacent devices during the migration process, and controlling the end-to-end transmission delay between nodes in the communication network to be lower than a preset maximum value, includes: Generate a device migration instruction including a migration path and a migration speed according to a relative distance between each target location in the target deployment location set and a current edge computing device location; Driving the edge computing device to migrate to the target location according to the device migration instruction, detecting the signal strength with the adjacent edge computing device during the migration process, and selecting the adjacent device with the signal strength higher than a preset threshold as a candidate node of the communication network; Based on the location distribution and signal strength data of the candidate nodes, a star-shaped communication network topology centered on the edge computing device is constructed; The end-to-end transmission delay between each topology node in the star communication network topology is monitored. If it is detected that the end-to-end transmission delay exceeds a preset maximum value, the candidate node is reselected and the hop count distribution of the star communication network topology is adjusted.

[0013] Optionally, the adjusting the electromagnetic coupling parameter combination of each charging port in the multi-port wireless charging system so that the output power range of the high priority charging port is adapted to the fast charging requirement and the output power range of the low priority charging port is adapted to the conventional charging requirement includes: Extracting a spatial density parameter of the fast charging demand within a coverage area of ​​a high-priority charging port according to the spatial distribution of the charging ports with different priorities marked in the priority matching rule; Based on the spatial density parameter, a narrowband resonant frequency range and a small-pitch power transmission coil parameter combination are configured for each high-priority charging port, so that the output power range of the high-priority charging port covers the power range corresponding to the fast charging demand; Synchronously configuring a broadband resonant frequency range and a large-pitch power transmission coil parameter combination for a low-priority charging port, so that the output power range of the low-priority charging port covers the power range corresponding to the conventional charging demand; The actual output power fluctuation data of the charging ports with different priorities are collected, and if a deviation from the corresponding power interval is detected, the resonant frequency range and the power transmission coil spacing parameter combination are recalculated based on the current fluctuation data.

[0014] In a second aspect, the present application provides a charging pile intelligent site selection and evaluation system, comprising: An acquisition module, used to acquire mobile terminal communication data of a target area, and generate a demand period division model based on the spatiotemporal variation law of the flow density of people in the mobile terminal communication data, wherein the demand period division model is used to distinguish between the fast charging demand corresponding to the daytime period and the conventional charging demand corresponding to the nighttime period; An adjustment module is used to adjust the position of the edge computing device deployed in the target area according to the demand distribution result output by the demand period division model, so that the service range of the edge computing device moves consistently with the peak area of ​​the pedestrian flow density and meets the preset data transmission performance requirements; a control module, configured to control the multi-port wireless charging system to switch a charging power output mode based on the charging demand type identified in the position adjustment result, wherein the charging power output mode corresponds to the time periods of the fast charging demand and the regular charging demand; A generation module, configured to combine the network layout data after the position adjustment of the edge computing device and the power output mode information of the multi-port wireless charging system to generate a set of candidate charging pile layout locations that simultaneously meet the daytime charging service efficiency and nighttime equipment utilization rate optimization indicators; The screening module is used to verify the underground pipe network safety distance for each location in the candidate charging pile layout location set, and screen out the target charging pile layout plan that meets the preset construction safety standards and matches the demand distribution law of the demand period division model.

[0015] In a third aspect, an embodiment of the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a charging pile intelligent site selection and evaluation method as described in the first aspect above.

[0016] In an embodiment of the present application, mobile terminal communication data of a target area is obtained, and a demand period division model is generated based on the spatiotemporal variation law of the crowd density in the mobile terminal communication data, wherein the demand period division model is used to distinguish between the fast charging demand corresponding to the daytime period and the conventional charging demand corresponding to the nighttime period; according to the demand distribution result output by the demand period division model, the position of the edge computing device deployed in the target area is adjusted so that the service range of the edge computing device moves in a consistent manner with the peak area of ​​the crowd density; based on the charging demand type identified in the position adjustment result, the multi-port wireless charging system is controlled to switch the charging power output mode, and the charging power output mode corresponds to the time period of the fast charging demand and the conventional charging demand; the network layout data after the position adjustment of the edge computing device is combined with the information corresponding to the power output mode of the multi-port wireless charging system, to generate a set of candidate charging pile layout positions; the distance verification is performed on each position in the set of candidate charging pile layout positions, and the target charging pile layout scheme that meets the preset safety standards and matches the demand distribution law of the demand period division model is screened out.

[0017] The technical solution of this application has the following beneficial effects: A dynamic demand model is built based on the temporal and spatial changes in real-time pedestrian density to accurately distinguish between daytime fast charging and nighttime regular charging needs, thereby improving the accuracy of demand perception and time period division. Real-time migration of edge device locations is used to match pedestrian peak areas to ensure dynamic alignment of computing resources with high-demand areas, reduce data transmission delays, and improve service response efficiency. Differentiated charging power output is configured according to demand types to achieve adaptive switching between fast charging and regular charging modes, solving the problems of equipment vacancy and grid load imbalance. Combined with equipment layout and power mode data, spatial candidate areas that meet both daytime efficiency and nighttime utilization are screened out to avoid resource waste caused by a single indicator. By matching safety distance constraints with demand distribution patterns, underground pipeline conflict areas are eliminated and equipment spacing is optimized to ensure the safety and demand adaptability of the deployment plan.

[0018] Furthermore, based on the network layout data adjusted by the edge computing device, the daytime high-efficiency area containing the peak point of the traffic density and the minimum service radius is identified; the nighttime high-utilization output area of ​​the wireless charging system (the overlapping range of the coil coverage and the demand hotspot) is obtained simultaneously, and the candidate layout range covering both the daytime efficiency boundary and the nighttime core area is screened out through geographic space superposition; the candidate areas are prioritized in combination with the charging pile spacing rules to generate a set of layout locations that take into account dynamic demand adaptation and engineering feasibility. Through spatial superposition and priority sorting, the coordinated optimization of daytime charging efficiency and nighttime equipment utilization is achieved, while avoiding underground pipe network conflicts and equipment overcrowding problems, and improving the dynamic adaptability and long-term operation and maintenance stability of the charging pile layout solution in mixed functional areas.

[0019] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 A flowchart of a charging pile intelligent site selection and evaluation method provided by the present application is shown; Figure 2 A schematic diagram of the structure of a charging pile intelligent site selection and evaluation system provided by the present application is shown; Figure 3 A schematic diagram of the structure of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0023] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0024] Researchers have found that the current charging pile site selection method has significant shortcomings in dynamic demand response and resource collaborative optimization. Especially for urban mixed-functional areas where the charging demand differs significantly between daytime and nighttime, the existing schemes are difficult to track the fluctuations in passenger density in real time, resulting in rigid charging pile layout, unbalanced equipment utilization, and intensified contradictions between peak and valley loads in the power grid. Based on this, a charging pile intelligent site selection and evaluation method is provided. Specifically, a spatiotemporal demand partitioning model is constructed through mobile terminal communication data, and the edge computing device migration is dynamically driven to match the peak passenger flow area. The power mode of the wireless charging system is adjusted based on the real-time demand type, and finally a safe and adaptive layout plan is generated by combining the equipment layout and power characteristics. This method can achieve high-precision matching of charging resources and dynamic demand, simultaneously optimize the daytime fast charging efficiency and the nighttime equipment utilization rate, and effectively avoid construction safety hazards.

[0025] The technical solution of the present application can be applied to scenarios where the difference in charging demand between daytime office work and nighttime living is balanced in urban mixed-functional areas (such as commercial and residential complexes and subway-covered properties).

[0026] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0027] Figure 1 A flowchart of a charging pile intelligent site selection and evaluation method is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes: 101. Obtain mobile terminal communication data of a target area, and generate a demand period division model based on the spatiotemporal variation of the crowd density in the mobile terminal communication data, wherein the demand period division model is used to distinguish between the fast charging demand corresponding to the daytime period and the conventional charging demand corresponding to the nighttime period; In this step, the demand period division model is a decision model constructed based on the spatiotemporal variation of the crowd density in the mobile terminal communication data, and is used to divide the target area into the daytime fast charging demand-dominated period and the nighttime conventional charging demand-dominated period.

[0028] In the embodiment of the present application, anonymous signaling data (including terminal device location, dwell time, and movement trajectory) collected by the communication base station is used to count the density of people flow in each grid unit in units of time windows. A hierarchical clustering algorithm is used to spatially aggregate high-density areas, and adjacent time periods with the same fluctuation pattern are identified in combination with time series similarity analysis. The demand switching nodes between daytime and nighttime are modeled through the state transition probability algorithm, and finally a demand period division model containing time period boundaries and demand intensity labels is generated.

[0029] Assume that the mobile terminal signaling data of the garage on the B1 floor of a commercial and residential complex shows that during the period of 8:00-10:00 on weekdays, the number of active terminals increased from 500 to 1,800 every 15 minutes, and the median stay time increased from 3 minutes to 25 minutes; while the number of active terminals during the period of 20:00-22:00 remained stable at around 300, and the median stay time reached 120 minutes. The demand period division model marks 8:00-20:00 as the "daytime fast charging demand period" (people flow density>1.2 people / ㎡), and 20:00-8:00 as the "nighttime regular charging demand period" (density<0.5 people / ㎡).

[0030] 102. According to the demand distribution result output by the demand time segmentation model, adjust the position of the edge computing device deployed in the target area so that the service range of the edge computing device moves in line with the peak area of ​​the crowd density; In this step, the edge computing device is a distributed computing node deployed in the target area, which has real-time data processing and communication capabilities and is used to dynamically adjust the service scope to adapt to the spatial movement of the peak pedestrian density area.

[0031] In the embodiment of the present application, based on the dynamic heat map output by the demand period division model, the moving trajectory coordinate sequence of the peak area of ​​crowd density is extracted. The multi-objective optimization problem is constructed by covering the radius constraint and the network delay constraint, and the device migration path is solved by the heuristic search algorithm. During the migration process, the signal strength and network topology status of adjacent devices are monitored in real time, and the shortest delay communication link is dynamically established to ensure that data interaction meets the real-time requirements. Finally, the optimal location set of edge computing devices and their corresponding service coverage are output.

[0032] For example, continuing with the above example, when the peak of the noontime traffic moves from the office area on the west side of the commercial and residential complex to the dining area on the east side, the edge computing device moves 120 meters along the preset path to the center of the new peak and establishes a double-hop communication link with the adjacent device. After the adjustment, the coverage rate of the traffic density in the device coverage area increased from 58% to 89%, and the end-to-end data transmission delay decreased from 68ms to 41ms.

[0033] 103. Based on the charging requirement type identified in the position adjustment result, control the multi-port wireless charging system to switch the charging power output mode, wherein the charging power output mode corresponds to the time period of the fast charging requirement and the regular charging requirement; In this step, the charging power output mode is a power configuration scheme that the multi-port wireless charging system dynamically adjusts according to the demand type, including a high-power fast charging mode and a low-power normal mode, and achieves differentiated output through electromagnetic coupling parameter control.

[0034] In an embodiment of the present application, based on the distribution of demand types identified in the coverage area of ​​the edge device, a priority weight is assigned to each charging port to switch the charging power output mode: the coil operating frequency band is adjusted through a resonant frequency adaptive matching algorithm (such as 6.78MHz narrowband mode for fast charging, 13.56MHz broadband mode for conventional charging), while optimizing the physical spacing of the power transmission coils (10cm tight coupling is used in fast charging mode, and 30cm loose coupling is used in conventional mode), and an impedance matching network is used to compensate in real time for energy loss caused by device displacement to ensure power output stability.

[0035] For example, continuing with the above example, the six charging ports in the dining area on the east side of the commercial and residential complex detected that the fast charging demand accounted for 82%, locked the resonant frequency at 6.78MHz, adjusted the coil spacing to 10cm, and increased the output power to 120kW; while the ports in the residential area on the north side switched to 13.56MHz broadband mode, the spacing was expanded to 30cm, and the power was stabilized at 22kW.

[0036] 104. Generate a set of candidate charging pile layout locations by combining the network layout data after the position of the edge computing device is adjusted with information corresponding to the power output mode of the multi-port wireless charging system; In this step, the candidate charging pile layout location set is a set of candidate locations generated by combining the equipment layout data and the power mode information, and must meet both the daytime service efficiency and nighttime equipment utilization indicators.

[0037] In the embodiment of the present application, the coverage of the edge computing device is discretized into high-precision geographic grids, and the daytime service efficiency score of each grid is calculated (based on crowd density, charging waiting time, and equipment load rate). The nighttime power output curve of the wireless charging system is simultaneously analyzed to screen the output areas with utilization rates higher than the average level. Through spatial superposition and conflict detection, prohibited construction areas (such as fire passages and green belts) are excluded, and finally a set of candidate charging pile layout locations covering daytime high-efficiency areas and nighttime high-utilization areas is generated.

[0038] For example, continuing with the above example, 35 candidate grids were selected on the southeast side of the commercial and residential complex, of which 12 grids met both the daytime performance score>8.0 (full score 10) and the nighttime utilization rate>70%. After excluding the road-occupying areas, 28 locations were retained to form the candidate set, and the distance between each location was ≥8 meters and ≤25 meters.

[0039] 105. Perform distance verification on each location in the candidate charging pile layout location set, and screen out a target charging pile layout solution that meets preset safety standards and matches the demand distribution law of the demand period division model.

[0040] In this step, the underground pipeline safety distance verification is a verification step to detect whether the straight-line distance between the candidate location and the underground gas pipeline and power pipeline corridor meets the construction safety standards.

[0041] In the embodiment of the present application, the three-dimensional coordinate data of the underground pipe network is converted into a spatial risk layer, and the Euclidean distance between each candidate location and the nearest risk source (gas pipe, cable well) is calculated. Penalty weights are applied to candidate locations whose distance is less than the safety threshold, and a comprehensive score is given in combination with the matching degree of demand distribution rules (such as the similarity of historical traffic density curves). The final target charging pile layout plan is screened out through multi-dimensional weighted sorting.

[0042] For example, continuing with the above example, of the 28 candidate locations, 9 locations were eliminated because they were less than 2 meters away from the gas pipeline, and 6 locations had a demand matching degree below the standard. Finally, 13 locations were selected to form the target charging pile layout plan, of which 8 were located in the core area of ​​the catering area and 5 were near the entrances and exits of the residential area.

[0043] Steps 101-105 achieve full-cycle intelligent decision-making for the layout of charging piles in mixed-functional urban areas by dynamically sensing changes in crowd density, adjusting edge computing resources in real time, adaptively optimizing charging power output, collaboratively screening candidate locations, and strictly verifying safety. In complex scenarios such as commercial and residential complexes, the response speed of daytime charging services and the utilization rate of equipment resources at night are significantly improved, construction safety hazards are effectively avoided, and a closed-loop optimization system that is demand-driven, resource-coordinated, safe and controllable is formed.

[0044] In order to further improve the coordinated optimization capability of the charging pile layout scheme between daytime efficiency and nighttime utilization, and solve the problems of rigid spatial resource allocation and split time demand response in the prior art, this application constructs a candidate location generation mechanism with spatiotemporal efficiency coupling by integrating edge computing dynamic layout data and wireless charging system power characteristics. In some embodiments, the network layout data after the position adjustment of the edge computing device and the power output mode information of the multi-port wireless charging system are combined to generate a set of candidate charging pile layout locations, including: 201. Extracting the network layout data after the position adjustment of the edge computing device, and identifying a continuous spatial area that meets the daytime charging service efficiency requirements; In step 201, the continuous space area refers to a set of physical spaces within the coverage of the edge computing device that meet the daytime charging service efficiency requirements, and must include both the peak point of pedestrian density and the minimum service radius constraint.

[0045] In an embodiment of the present application, based on the network topology data after the migration of the edge computing device, an improved spatial clustering algorithm (such as a density-based DBSCAN variant) is used to rasterize the coverage area. First, the device service radius is discretized into 50cm×50cm grid units, and the daytime charging service efficiency index of each grid is calculated (compositely generated by the weighted value of the flow density, the charging request response time attenuation coefficient, and the load balancing degree of the equipment). Discrete noise points are eliminated through a multi-scale morphological filtering algorithm, and high-efficiency grid clusters with spatial continuity are retained. Finally, the regional growing algorithm is used to merge adjacent high-efficiency grids to form a continuous spatial area that meets the minimum service area threshold.

[0046] 202. Synchronously obtain power output mode information of the multi-port wireless charging system, and mark output areas where the power utilization rate during the night time period is continuously higher than the average level; In step 202, the output area is the effective coverage of the multi-port wireless charging system where the power utilization rate is continuously higher than the average level during the night time period, which is the spatial intersection of the electromagnetic field intensity distribution of the power transmission coil and the conventional charging demand hot zone.

[0047] In the embodiment of the present application, the historical power output curve of each port is collected by the wireless charging control module, and the stable high utilization period at night (such as 22:00-6:00) is extracted by the time series piecewise aggregation approximation (PAA) algorithm. A coil coverage intensity heat map is constructed based on the electromagnetic field finite element simulation data, and the spatial distribution of conventional charging demand hotspots (from the demand period division model) is superimposed. The spatial convolution operation is used to fuse the features of the heat map and the demand distribution map, and the grid set with electromagnetic field strength values ​​greater than the critical intensity and demand density higher than the regional mean is screened out, and marked as a high utilization output area.

[0048] 203. Performing geographic spatial superposition on the continuous spatial area and the output area, and screening out candidate layout ranges that include both the daytime efficiency area boundary and the nighttime utilization rate core area; In step 203, the candidate layout range is an optimization interval generated by spatially superimposing the continuous spatial area and the output area, and must include both the boundary constraints of the daytime efficiency area and the coverage of the nighttime utilization core area.

[0049] In the embodiment of the present application, the raster data of the continuous space area and the output area are converted into vector polygons, and the overlapping area is obtained by a topological overlay algorithm (such as an intersection operation in a Boolean operation). The geometric contours of the daytime efficiency area are identified by a boundary extraction algorithm, and a spatial constraint network is constructed in combination with Delaunay Triangulation. Each sub-area in the overlapping area is scored with multiple indicators (daytime efficiency weight 60%, nighttime utilization weight 40%), and the non-dominated sorting genetic algorithm (NSGA-II) is used to screen the Pareto optimal solution set to generate a candidate layout range that takes into account the dual-time period indicators.

[0050] 204. Based on the candidate deployment range, spatial priority is sorted on the deployable areas according to a preset charging pile spacing rule to generate a set of candidate charging pile deployment locations.

[0051] In step 204, the spatial priority sorting is to perform a hierarchical evaluation of the deployable areas within the candidate layout range based on the preset charging pile spacing rules, and give priority to the locations with the best overall performance and meeting the engineering constraints.

[0052] In an embodiment of the present application, a spatial optimization model with constraints is constructed within the candidate layout range. First, an improved ant colony algorithm (with the introduction of a dynamic pheromone volatilization mechanism) is used to traverse all feasible locations, and the neighboring charging pile coverage redundancy of each location is calculated (the inverse weighted sum of the distances between adjacent locations). The geographic accessibility of the location (generated based on the path planning algorithm to generate the shortest arrival path length) and the grid access cost (converted based on the distance to the substation facility) are simultaneously evaluated. The above indicators are jointly optimized through the multi-objective particle swarm optimization algorithm (MOPSO), and a set of candidate charging pile layout locations arranged in descending order by comprehensive score is output.

[0053] Here is a specific example: In the underground parking lot of a certain property complex above a subway, the peak flow density in the west office area during the daytime commuting period (7:00-19:00) reaches 2.1 people / m2, and the charging demand in the east residential area during the nighttime residential period (20:00-6:00) accounts for more than 75%. Step 201 extracts the coverage data of the edge computing equipment after migration, and identifies the continuous space area on the west side (area 1200㎡, including 6 peak points of traffic flow, service radius ≥15 meters); step 202 synchronously obtains the wireless charging system data, and marks the high-utilization output area on the east side at night (electromagnetic field strength>3A / m, demand density>0.8 people / ㎡); through step 203, the candidate layout range of the north-south transition zone is screened out after spatial superposition (area 800㎡, daytime efficiency score>7.5, nighttime utilization>65%); step 204 generates 32 candidate locations by sorting according to the spacing rule, of which 18 are located in the core corridor area (spacing 12-20 meters, grid access distance <50 meters), and generates a set of candidate charging pile layout locations.

[0054] Steps 201-204 achieve a dynamic balance between daytime efficiency and nighttime utilization of charging pile layout through the coordination of spatiotemporal clustering, electromagnetic field fusion, topological superposition and multi-objective optimization technologies. In the scenario of urban mixed-functional areas, the spatiotemporal adaptability of charging resources is significantly improved, the equipment idle rate and the risk of power grid fluctuations are reduced, and the full-process optimization capability from data fusion, spatial analysis to project implementation is formed.

[0055] In order to further improve the safety of the charging pile layout plan and the matching accuracy of the demand pattern, and solve the problems of separation between safety verification and dynamic demand adaptation and high manual correction costs in the prior art, this application integrates the three-dimensional data of the underground pipeline network, the spatial distribution of demand hotspots, and the historical fluctuation pattern of passenger flow to construct a multi-dimensional linkage safety screening and dynamic adaptation mechanism. In some embodiments, the distance verification is performed on each location in the candidate charging pile layout location set to screen out the target charging pile layout plan that meets the preset safety standards and matches the demand distribution pattern of the demand period division model, including: 301. Obtain the spatial distance between each location in the candidate charging pile layout location set and the three-dimensional coordinate data of the underground pipe network, and select a subset of each location whose straight-line distance to the gas pipeline and the power pipeline corridor is greater than a preset safety distance; In step 301, the preset safety distance is the minimum straight-line distance threshold between the preset charging pile layout location and infrastructure such as underground gas pipelines and power pipelines, and must comply with municipal construction safety regulations.

[0056] In the embodiment of the present application, the three-dimensional point cloud data of the underground pipeline network is loaded through the three-dimensional geographic information system (GIS) engine, and the candidate position coordinates are converted into a spatial rectangular coordinate system. The improved K-Nearest Neighbors (K=5) algorithm is used to retrieve the nearest gas pipeline node and the center point of the power corridor section around each location, and the minimum Euclidean distance is calculated. The Monte Carlo risk simulation method is introduced to estimate the probability density of the distance distribution, eliminate the positions with a safety probability lower than the confidence interval (such as 95%), and finally generate a subset containing all the safe distance standard positions.

[0057] 302. Based on the spatial distribution of the fast charging demand hotspots and the conventional charging demand hotspots marked in the demand period division model, calculate the area ratio of the subset covering the two types of demand hotspots, and retain the adaptation position that meets the lower limit of the coverage of the daytime fast demand and the nighttime conventional demand while retaining the area ratio; In step 302, the area ratio is the ratio of the spatial coverage area of ​​the candidate location subset to the fast charging demand hot zone and the conventional charging demand hot zone, which must meet the coverage lower limit thresholds both during the day and at night.

[0058] In the embodiment of the present application, the vector boundary data of the fast charging demand hot zone and the conventional charging demand hot zone are rasterized (resolution 0.5m×0.5m), and the region growing algorithm (Region Growing) is used to perform spatial overlay analysis on the candidate location subset. The number of grids covering the two types of hot zones in each location buffer is counted, and the area ratio is calculated (fast demand coverage ratio = number of grids covering hot zones / total number of grids). The lower limits of daytime and nighttime coverage are dynamically set through an adaptive threshold segmentation algorithm (Otsu variant), and positions that meet both thresholds are retained. The remaining positions are tested for spatial autocorrelation (Moran's I index) to remove isolated points.

[0059] 303. Verify the underground pipe network safety spacing of the adaptation position, remove abnormal positions whose spacing with adjacent pipe network positions exceeds a preset spacing interval, and generate a set of intermediate layout solutions; In step 303, the intermediate layout plan set is a temporary layout plan set generated after multi-level safety distance verification, and the final plan needs to be further screened in combination with dynamic demand rules.

[0060] In the embodiment of the present application, based on the three-dimensional pipe network topology relationship model, an implicit surface (ImplicitSurface) is constructed to express the safety envelope of the pipe gallery. The ray casting algorithm (Ray Casting) is used to calculate the shortest distance from the adaptation position to the envelope surface. If the distance is less than the minimum spacing or greater than the maximum spacing, it is marked as abnormal. The hidden Markov model (HMM) is introduced to analyze the spatial aggregation pattern of abnormal positions and identify systemic risk areas (such as pipeline intersections). The final output is a set of intermediate layout solutions constrained by multi-level spacing.

[0061] 304. Perform safety assessment based on the historical time period passenger density curve in the intermediate layout plan set, perform similarity matching between the test result and the demand distribution law of the demand period division model, and screen out the target charging pile layout plan that meets the preset safety standards and matches the demand distribution law.

[0062] In step 304, the safety assessment combines the historical passenger density curve and demand distribution law to verify the matching degree of the intermediate solution with the dynamic demand fluctuation.

[0063] In the embodiment of the present application, the historical traffic density time series (sampling frequency 15 minutes) of each location in the intermediate solution is extracted, and the dynamic time warping (DTW) algorithm is used to calculate the similarity distance with the benchmark curve of the demand period division model. The similarity distance matrix is ​​divided into high-matching clusters and low-matching clusters through spectral clustering. The spatial kernel density estimation (KernelDensity Estimation) is performed on the locations in the high-matching clusters, and the density peak points are screened to generate the target charging pile layout plan.

[0064] Here is a specific example: In the underground parking lot of a subway-covered property complex, the office area on the west side forms a fast charging demand hot zone (area 600 m2) during the day (7:00-19:00), and the residential area on the east side forms a conventional demand hot zone (area 450 m2) at night (20:00-6:00). The candidate layout location set contains 50 points. After screening in step 301, 32 subsets containing all safe distance-compliant locations are retained (distance to gas pipelines>2.5m, distance to power corridors>1.8m). The area ratio calculation in step 302 shows that 18 of the locations simultaneously cover ≥55% of the daytime hot zone area and ≥40% of the nighttime hot zone area. Step 303 further eliminates 4 abnormal spacing locations (near drainage wells, spacing <0.8m), generating 14 intermediate layout plan sets. In step 304, based on the six-month historical data, 8 target locations (DTW similarity>0.75, kernel density peak>1.2) are screened and generated, distributed in the west channel and the east transition zone, with a spacing of 12-22 meters, to generate a target charging pile layout plan.

[0065] Steps 301-304 achieve a precise balance between underground pipe network safety constraints and dynamic demand adaptation in the charging pile layout plan through three-dimensional space distance screening, dynamic area ratio calculation, multi-level safety distance verification, and historical law matching evaluation. In the scenario of urban mixed-functional areas, construction safety and demand response consistency are significantly improved, operation and maintenance risks and resource mismatch rates are reduced, and a full-link closed-loop decision-making capability is formed from geometric constraints, spatial analysis to time series matching.

[0066] In order to further improve the dynamic response capability and network transmission stability of edge computing device deployment, and solve the problem of insufficient real-time demand response caused by high device migration lag and rigid network topology in the prior art, this application constructs an edge computing resource scheduling system for adaptive peak flow movement through spatiotemporal trajectory prediction, dynamic location optimization, network reconstruction and closed-loop feedback mechanism. In some embodiments, the position of the edge computing device deployed in the target area is adjusted according to the demand distribution result output by the demand period division model, including: 401. Analyze the demand distribution result output by the demand period division model, extract the high demand area coordinate set in the target area in the current period, and determine the moving trajectory of the peak area of ​​the crowd density according to the spatiotemporal distribution characteristics of the high demand area coordinate set; In step 401, the high-demand area coordinate set is a set of geographical locations in the target area where the density of people flows exceeds a preset threshold in the current time period, and is used to characterize the charging demand concentration area.

[0067] In the embodiment of the present application, a multi-scale analysis is performed on the dynamic distribution results output by the demand period division model through a spatiotemporal density clustering algorithm (such as an improved OPTICS algorithm) to extract the core clusters that meet the density threshold. The time series piecewise linear representation (PLR) algorithm is used to compress the spatiotemporal coordinate sequence and identify the drift direction of the center of gravity of the clusters in adjacent time periods. Based on the hidden Markov model (HMM), a state transition probability matrix is ​​established, the continuous movement path of the peak area is decoded, and a movement trajectory containing timestamps and spatial coordinates is generated.

[0068] 402. Calculate a target deployment position set of the edge computing device based on the center position of the peak area of ​​each time segment in the moving trajectory, wherein the straight-line distance between each target position in the target deployment position set and the center of the peak area of ​​the corresponding time segment does not exceed a preset coverage radius; In step 402, the target deployment location set is the optimal location set to which the edge computing device needs to migrate, ensuring that the distance between each location and the center of the peak area of ​​the corresponding time period does not exceed a preset coverage radius.

[0069] In an embodiment of the present application, a reinforcement learning framework (such as Q-Learning) is used to construct a location optimization model based on the spatiotemporal coordinate points in the mobile trajectory sequence. The state space is defined as the Euclidean distance difference between the real-time coordinates of the peak area and the current position of the device, and the action space is the device migration direction (8-neighborhood discretization) and step size (50m / 100m). The reward function integrates the coverage gain (incremental density of pedestrian flow within the service range) and the migration cost (energy consumption + network reconstruction overhead). The Q value table is iteratively updated through the ε-greedy strategy to output the target deployment location set that meets the coverage radius constraint and maximizes the cumulative reward.

[0070] 403. Generate a device migration instruction according to the target deployment location set, drive the edge computing device to migrate to the target location according to the device migration instruction, and establish a communication network with adjacent devices during the migration process, and control the end-to-end transmission delay between nodes in the communication network to be lower than a preset maximum value; In step 403, the device migration instruction is a control instruction set that drives the edge computing device to migrate to the target location along a preset path, including a migration path, speed, and network reconstruction parameters.

[0071] In the embodiment of the present application, the target deployment location set is input into the path planning algorithm (such as the improved A* algorithm) to generate device migration instructions based on the optimal migration path to avoid obstacles (avoiding fixed obstacles such as fire passages and load-bearing columns). During the migration process, the communication link establishment priority is negotiated with adjacent devices through a distributed consensus protocol (Raft algorithm variant) to dynamically build a minimum spanning tree (MST) network topology. The link state routing protocol (OSPF improved version) is used to calculate the shortest transmission path between nodes in real time, and the single-hop delay is limited through traffic shaping technology (Token Bucket algorithm) to ensure that the end-to-end delay is lower than the preset maximum value.

[0072] 404. Obtain an updated coordinate set of the peak area through the communication network. When it is detected that the coverage overlap rate between the updated coordinate set and the target deployment location set is lower than a preset tolerance threshold, re-execute the target deployment location set calculation based on the updated coordinate set, and synchronously adjust the device migration instruction.

[0073] In step 404, the coverage overlap ratio is the ratio of the intersection area of ​​the coverage area of ​​the updated coordinate set and the target deployment position set to the coverage area of ​​the original position set, which is used to quantify the degree of trajectory deviation.

[0074] In the embodiment of the present application, the updated coordinate set of the peak area is collected in real time through the communication network, and the Jaccard similarity coefficient is used to calculate the coverage overlap rate between it and the target position set. If the overlap rate is lower than the tolerance threshold, the incremental trajectory prediction process is triggered: the mobile trajectory prediction error is corrected based on the Kalman filter algorithm, and steps 401-402 are re-executed to generate a new target deployment position set. The Receding Horizon Control strategy is used to dynamically adjust the migration instructions to ensure a smooth transition of the device migration path and minimize the network topology reconstruction overhead.

[0075] Here is a specific example: In the underground parking lot of a subway-covered property complex, the peak passenger flow area during the day (7:00-19:00) moves periodically from the west security checkpoint to the central commercial area according to the commuter tide. Step 401 parses the demand distribution data, extracts the high-demand coordinate set (west coordinate cluster A, density > 2 people / m2) during the peak period (8:00-9:30), and generates a mobile trajectory (drifting west → northeast → east) through HMM decoding; Step 402 calculates the target location set (locations P1-P5) through the Q-Learning model, and sets the coverage radius to 30 meters to ensure that the distance between each location and the center point of the trajectory is ≤25 meters; Step 403 generates a migration path (west P1 → northeast P3 → east P5) through the A* algorithm, and establishes a three-hop MST network with adjacent devices during the migration process, with the end-to-end delay stabilized below 45ms; In step 404, it is detected that the updated coordinate set moves 200 meters east at noon (due to temporary commercial activities), and the coverage overlap rate drops to 40%, triggering the trajectory correction and regenerating the target deployment location set (adding locations P6-P8), and the overlap rate recovers to 82% after the device migrates to the new location.

[0076] Steps 401-404 achieve real-time tracking and adaptive coverage of peak traffic areas by edge computing devices through the collaboration of spatiotemporal clustering, reinforcement learning, distributed network optimization, and rolling time domain control technologies. In the scenario of mixed-function areas in cities, the timeliness of charging demand response is significantly improved, network transmission delay and device migration energy consumption are reduced, and closed-loop optimization capabilities from trajectory prediction, resource scheduling to dynamic correction are formed to ensure service continuity in complex scenarios.

[0077] In order to further improve the power allocation accuracy and adaptability of the wireless charging system in dynamic demand scenarios, and solve the problems of power mode switching lag, equipment idleness and overload in the prior art, this application constructs a demand-driven wireless charging power control system through demand type identification, priority dynamic configuration, electromagnetic parameter optimization and closed-loop feedback mechanism. In some embodiments, the control of switching the charging power output mode of the multi-port wireless charging system based on the charging demand type identified in the position adjustment result includes: 501. Obtain a charging demand type marked in a position adjustment result, where the charging demand type includes a first demand signal and a second demand signal corresponding to the fast charging demand and the conventional charging demand, respectively; In step 501, the charging demand type is a category of charging demand marked according to a demand period classification model, including a first type of demand signal corresponding to a fast charging demand and a second type of demand signal corresponding to a conventional charging demand.

[0078] In the embodiment of the present application, the charging demand type label marked in the spatiotemporal label matrix is ​​extracted through the location adjustment result data uploaded by the edge computing device. The spatiotemporal clustering algorithm (such as the improved ST-DBSCAN) is used to spatially aggregate the demand signal to separate the high-density fast charging demand cluster (first type of signal) and the low-density conventional demand cluster (second type of signal), and the environmental noise interference is eliminated based on the blind source separation technology (BSS).

[0079] 502. According to the spatial distribution density of the charging demand type, configure a priority matching rule for each charging port of the multi-port wireless charging system, wherein the priority of the charging port in the area covered by the first type of demand signal is higher than that of the second type of demand signal; In step 502, the priority matching rule is a service priority strategy assigned to the wireless charging port, ensuring that the port priority of the first type demand signal coverage area is higher than that of the second type demand signal area.

[0080] In the embodiment of the present application, based on the spatial distribution density heat map of the demand type, a spatial convolution kernel (Gaussian Kernel) is used to calculate the integral value of the demand density within the coverage area of ​​each charging port. A multi-port collaborative optimization model is constructed through the dynamic game theory framework (Nash Bargaining Solution), and the port utility function is defined as the weighted difference between the integral value of the demand density and the interference coefficient of the adjacent port. The alternating direction multiplier method (ADMM) is used to solve the Nash equilibrium point and generate a priority weight matrix for configuring the priority matching rules, where the priority weight coefficient of the port in the first type of signal area is not less than 1.5 times that of the second type of area.

[0081] 503. Based on the priority matching rule, adjust the electromagnetic coupling parameter combination of each charging port in the multi-port wireless charging system so that the output power range of the high-priority charging port adapts to the fast charging requirement, and the output power range of the low-priority charging port adapts to the conventional charging requirement, wherein the electromagnetic coupling parameter combination includes a resonant frequency matching range and a power transmission coil spacing; In step 503, the electromagnetic coupling parameters are combined into a set of operating parameters of each port in the multi-port wireless charging system, including a resonant frequency matching range and a power transmission coil spacing, which are used to control the output power mode.

[0082] In the embodiment of the present application, the electromagnetic coupling parameter combination is adjusted according to the priority matching rule, and a narrowband resonant frequency (such as 6.78MHz±0.5MHz) and a small pitch coil layout (≤15cm) are configured for the high priority port, and a dynamic impedance matching network (based on real-time calibration of the Smith chart) is used to optimize the energy transmission efficiency. The low priority port is switched to a broadband resonant frequency (13.56MHz±2MHz) and a large pitch coil layout (≥30cm), and the electromagnetic field uniformity of the coil array is optimized by a genetic algorithm (NSGA-II), and the parameter adaptation surface is constructed using finite element simulation data to obtain the Pareto optimal solution.

[0083] 504. Periodically detect the degree of compatibility between the actual output power of each charging port and the priority matching rule through the electromagnetic coupling state feedback data of the multi-port wireless charging system. If it is detected that the degree of compatibility is lower than the matching rule requirement, regenerate the electromagnetic coupling parameter combination and update the priority matching rule.

[0084] In step 504, the adaptability is the degree of conformity between the actual output power of the charging port and the target power range required by the priority matching rule, which is used to quantify the effectiveness of the parameter configuration.

[0085] In the embodiment of the present application, the real-time output power, efficiency and coil temperature data of each port are collected through the wireless charging control module, and the deviation between the actual output and the target power is calculated by covariance analysis (ANCOVA). If the deviation exceeds the preset threshold, the parameter re-optimization process is triggered: based on the Kalman filter algorithm, the demand density change trend is predicted, and steps 502-503 are re-executed to generate updated priority rules and parameter combinations. The receding horizon optimization strategy is used to achieve dynamic parameter fine-tuning to ensure continuous adaptation to demand fluctuations.

[0086] Here is a specific example: The underground parking lot of a commercial and residential complex is divided into the west office area (intensive fast charging demand during the day) and the east residential area (dominated by conventional demand at night). In step 501, after the edge device is migrated, the west coordinates (X1, Y1) to (X5, Y5) are marked as the first type of demand signal (density>1.8 people / ㎡), and the east coordinates (X6, Y6) to (X10, Y10) are marked as the second type of signal (density<0.6 people / ㎡); the Gaussian kernel convolution calculation in step 502 shows that the demand density integral value of the west port P1-P3 is 4.2, and the east port P4-P6 is 1.5. The game model assigns a priority weight of 2.0 for P1-P3 and 2.0 for P4-P6. P6 is 1.3; in step 503, P1-P3 are configured with narrowband resonance of 6.78MHz and coil spacing of 12cm, and the output power is increased to 120kW; P4-P6 are switched to broadband of 13.56MHz and spacing of 35cm, and the power is stabilized at 25kW; in step 504, it is detected at noon that the actual output power of P2 is only 95kW (target 120kW), and the covariance analysis shows that the adaptability is insufficient. After re-optimization, its resonant frequency is adjusted to 6.80MHz, and the power is restored to 118kW.

[0087] Steps 501-504 achieve high-precision adaptation of wireless charging power output and dynamic demand through demand signal separation, game optimization, dynamic matching of electromagnetic parameters and closed-loop feedback control. In the scenario of urban mixed-functional areas, the response speed of fast charging demand and the utilization rate of conventional charging equipment are significantly improved, energy loss and equipment overheating risks are reduced, and full-link adaptive capabilities are formed from demand identification, rule configuration to parameter closed-loop optimization.

[0088] In order to improve the spatiotemporal accuracy and dynamic adaptability of the charging demand period division, and solve the problem of rigid period division caused by complex fluctuations in the flow of people in the existing model in the mixed functional area scene, this application constructs a dynamically evolving demand period division system through spatiotemporal grid analysis, density pattern recognition, overlap optimization and model self-correction mechanism. In some embodiments, the demand period division model is generated based on the spatiotemporal variation law of the flow density in the mobile terminal communication data, including: 601. Extracting the spatial coordinates and timestamp of the terminal device in the mobile terminal communication data, and constructing a spatiotemporal distribution sequence of the crowd density with the geographic grid as a unit; In step 601, the spatiotemporal distribution sequence of the crowd density is a terminal device distribution density data sequence that is statistically analyzed by time window with geographic grids as units, and is used to quantify the aggregation characteristics of crowds in the spatiotemporal dimension.

[0089] In the embodiment of the present application, the anonymous signaling data collected by the mobile terminal communication base station is used to extract the device MAC address, timestamp and longitude and latitude coordinates. The time series segmented aggregation approximation (PAA) algorithm is used to divide the original data stream into equal-length time periods (such as 15 minutes), and the coordinate points in each time period are spatially gridded (the grid size dynamically adapts to the area of ​​the region). The sparse grid data is supplemented based on the Kriging Interpolation algorithm to generate a spatiotemporal distribution sequence of crowd density containing grid center coordinates, crowd density values ​​and time tags. The sequence data is updated through a sliding window mechanism to ensure real-time performance.

[0090] 602. Perform pattern recognition on the fluctuation characteristics of the crowd density in each geographic grid in the crowd density spatiotemporal distribution sequence, extract the grid set with short-term high-density aggregation characteristics in the daytime period as the candidate area for fast charging demand, and extract the grid set with continuous medium-low density distribution in the nighttime period as the candidate area for conventional charging demand; In step 602, the candidate area for fast charging demand is a set of grids where the density of people surges in the daytime and the duration of the gathering is less than a preset threshold. The candidate area for regular charging demand is a set of grids where the density of people remains stable at a medium-low level during the nighttime and the duration exceeds a preset threshold.

[0091] In the embodiment of the present application, for each grid unit in the spatiotemporal distribution sequence, a morphological clustering algorithm is used to identify the density fluctuation pattern. For the daytime period, the dynamic time warping (DTW) algorithm is used to match the short-term spike waveform (duration ≤ 2 hours, peak density ≥ 1.5 people / ㎡) to screen out the candidate areas with fast demand; for the nighttime period, the autoregressive integrated moving average model (ARIMA) is used to detect the stability index to screen out the candidate areas with continuous medium and low density (0.3-0.8 people / ㎡) and variance below the threshold. Discrete noise grids are eliminated through spatial connectivity analysis, and the vector boundary data of the candidate area is output.

[0092] 603. Based on the spatiotemporal overlap analysis between the candidate areas for fast charging demand and the candidate areas for conventional charging demand, the candidate areas whose spatial coverage areas in adjacent time segments exceed the upper limit of the overlap are merged into demand period division units, and a corresponding charging demand type label is marked for each division unit; In step 603, the demand period division unit is a continuous spatiotemporal block generated by merging candidate areas with spatiotemporal overlap, which is marked as a fast or regular charging demand dominant type.

[0093] In the embodiment of the present application, the spatiotemporal coordinates of the fast and conventional candidate areas are mapped to a four-dimensional tensor space (longitude, latitude, time start and end, density level), and the community discovery algorithm (Louvain Method) in graph theory is used to identify high-overlap candidate area clusters. The spatiotemporal overlap is defined as the intersection ratio of the spatial coverage area of ​​the candidate areas in adjacent time periods. If the overlap exceeds 75%, they are merged into demand period division units, and the dominant charging demand type is labeled for each unit through the label propagation algorithm (Label Propagation).

[0094] 604. According to the label distribution rule of the demand time division unit in the historical period, the division granularity of the geographic grid and the determination threshold of the spatiotemporal overlap analysis are modified to generate a demand time division model adapted to the changing characteristics of the flow of people in the target area.

[0095] In step 604, the demand period division model is a decision model generated after dynamically correcting the grid granularity and overlap threshold based on historical data, and is used to predict the demand period distribution of the target area.

[0096] In the embodiment of the present application, the label distribution data of the division unit in the historical period is extracted, and the grid granularity and overlap threshold are adjusted using the Bayesian Optimization framework. The prediction accuracy (F1-score) of the parameter combination is evaluated by cross-validation, and the Pareto optimal solution is selected to update the model parameters. The long short-term memory network (LSTM) is used to model the time series dependency of the modified parameters, and finally the demand period division model that can adapt to the dynamic changes of passenger flow is output.

[0097] Here is a specific example: The underground parking lot on the second floor of a subway property complex has obvious characteristics of daytime commuting (surge of people at the west entrance from 7:00 to 10:00) and nighttime residence (stable demand at the east residential area from 20:00 to 6:00). Step 601 divides the mobile signaling data into 15-minute windows and constructs a spatiotemporal sequence of 50m×50m grids. The density of the west grid G12 increases sharply from 0.5 people / ㎡ to 2.3 people / ㎡ from 8:00 to 9:00. In step 602, the MCL algorithm identifies G12 as a fast candidate area (peak duration). The eastern grid E7-E9 is marked as a regular candidate area (density 0.4-0.7 people / m2 for 8 hours). In step 603, the Louvain algorithm merges the adjacent time period G12-G15 candidate areas (overlap 83%) to generate the western division unit U1 (fast-dominated), and merges E7-E9 into unit U2 (regular-dominated). In step 604, the grid size is reduced to 30m×30m through Bayesian optimization, the overlap threshold is increased to 85%, and the accuracy of the LSTM prediction model is increased to 92%.

[0098] Steps 601-604 achieve high-precision dynamic modeling of demand time segmentation through spatiotemporal grid analysis, density pattern clustering, community discovery optimization, and parameter self-correction technology. In the scenario of urban mixed-functional areas, the adaptability of time segmentation to complex fluctuations in passenger flow is significantly improved, the resource mismatch rate is reduced, and the full-process optimization capability from data quantification, pattern mining to model self-evolution is formed.

[0099] In order to further improve the network stability and real-time performance during the migration of edge computing devices, and solve the problems of communication interruption, topology rigidity and delay fluctuation caused by device movement in the prior art, this application builds a highly reliable edge computing network through path planning optimization, dynamic node selection, network structure adaptive adjustment and delay closed-loop control mechanism. In some embodiments, the device migration instruction is generated according to the target deployment location set, the edge computing device is driven to migrate to the target location according to the device migration instruction, and a communication network is established with adjacent devices during the migration process, and the end-to-end transmission delay between nodes in the communication network is controlled to be lower than the preset maximum value, including: 701. Generate a device migration instruction including a migration path and a migration speed according to a relative distance between each target location in the target deployment location set and a current edge computing device location; In step 701, the device migration instruction is a control instruction set including the migration path and speed of the edge computing device, which is used to guide the device to migrate from the current location to the target location.

[0100] In the embodiment of the present application, a multi-objective genetic algorithm (NSGA-II) is used to optimize the migration path based on the relative distance matrix between the target position set and the current equipment position. The objective function is defined as the shortest path length and the maximum obstacle avoidance (avoiding fixed obstacles such as fire passages and structural columns), and the constraints include the maximum migration speed (≤1m / s) and the acceleration limit (≤0.5m / s²). The risk of sudden obstacles during the migration process is predicted by the dynamic programming algorithm (Dyna-Q), and an alternative detour path is obtained, and finally the equipment migration instruction including the main path, the alternative path and the segmented speed curve is output.

[0101] 702. Drive the edge computing device to migrate to the target location according to the device migration instruction, detect the signal strength with the adjacent edge computing device during the migration process, and select the adjacent device with the signal strength higher than a preset threshold as a candidate node of the communication network; In step 702, the candidate nodes are adjacent edge computing devices whose signal strength during the migration process is higher than a preset threshold, and are used to build a temporary communication network.

[0102] In the embodiment of the present application, during the device migration process, the signal strength (RSSI) and signal-to-noise ratio (SNR) of the surrounding devices are scanned in real time through the distributed radio frequency fingerprint acquisition module. The Kalman filter algorithm is used to eliminate environmental noise interference, and the signal strength threshold is dynamically updated (initial -70dBm, adaptively adjusted with the migration speed). The node stability is evaluated based on the fuzzy logic reasoning system (signal strength fluctuation rate <15%, packet loss rate <5%), and a list of high-reliability candidate nodes is screened out. The node availability is confirmed through a lightweight consensus protocol (Paxos variant) to generate candidate nodes for the communication network.

[0103] 703. Construct a star-shaped communication network topology centered on the edge computing device based on the location distribution and signal strength data of the candidate nodes; In step 703, the star-shaped communication network topology is a communication network structure that uses the migrating edge computing device as a central node and establishes a single-hop connection with the candidate node.

[0104] In the embodiment of the present application, based on the location distribution and signal strength data of the candidate nodes, the closeness centrality algorithm in graph theory is used to determine the central node to ensure that the maximum single-hop coverage radius is ≤50 meters. The minimum spanning tree (Prim algorithm) is used to optimize the connection weights between nodes (based on the inverse weighting of signal strength and physical distance) to build a low-latency backbone link. The time division multiple access (TDMA) protocol is used to allocate communication time slots to avoid channel conflicts. Finally, a star communication network topology containing the central node coordinates, connection relationships, and time slot tables is generated.

[0105] 704. Monitor the end-to-end transmission delay between each topology node in the star communication network topology. If it is detected that the end-to-end transmission delay exceeds a preset maximum value, reselect the candidate node and adjust the hop count distribution of the star communication network topology.

[0106] In step 704, the end-to-end transmission delay is the total time taken for data packet transmission between any two nodes in the star network topology, and needs to be continuously monitored and controlled within a preset maximum value.

[0107] In the embodiment of the present application, the transmission delay data of each node is periodically collected through the link state routing protocol (OLSR improved version), and the exponential weighted moving average (EWMA) algorithm is used to predict the delay trend. If the delay is detected to exceed the threshold (such as 100ms), the topology reconstruction process is triggered: based on the reinforcement learning framework (Q-Learning), the candidate nodes are reselected, and the devices with low load and high signal strength are given priority. The dynamic hop constraint algorithm (DHC) is used to adjust the network hierarchy, reduce the maximum number of hops from 2 hops to 1 hop, and limit burst traffic through traffic shaping technology (Leaky Bucket).

[0108] Here is a specific example: In an underground parking lot of a commercial and residential complex, the west office area needs to frequently migrate edge computing devices during the day to match the commuting tidal flow of people. In step 701, the target location set includes P1 (coordinates X=120, Y=80) and P2 (X=150, Y=100). NSGA-II constructs the main path and generates device migration instructions (avoiding the B2 load-bearing column). The speed curve is divided into 0.8m / s (straight segment) and 0.3m / s (turning segment). In step 702, the adjacent device E3 (RSS I=-68dBm), E5 (RSSI=-72dBm), fuzzy logic selects E3 as the candidate node; in step 703, a star communication network topology is constructed with P1 as the center, connecting E3 and E7 nodes (single hop distance ≤45 meters), and TDMA allocates time slots of 0.5ms / node; in step 704, the delay from E3 to E7 is detected to 110ms during the lunch rush hour, and Q-Learning switches to the backup node E9 (RSSI=-65dBm), the number of hops is adjusted to 1 hop, and the delay is reduced to 75ms.

[0109] Steps 701-704 achieve high network reliability and low latency during the migration of edge computing devices through multi-objective path planning, fuzzy node selection, centrality network construction and delay-driven topology optimization. In the scenario of urban mixed-functional areas, the communication continuity in dynamic environments is significantly improved, the risk of data transmission interruption is reduced, and a full-link closed-loop management capability is formed from path control, node optimization to network self-healing.

[0110] In order to further improve the power output stability and energy efficiency adaptability of the wireless charging system in dynamic demand scenarios, and solve the problems of rigid power configuration, equipment overload and vacancy in the prior art, this application constructs a demand-driven wireless charging power control system through spatial density quantification, parameter differentiation configuration, dynamic tuning and closed-loop feedback mechanism. In some embodiments, the electromagnetic coupling parameter combination of each charging port in the multi-port wireless charging system is adjusted so that the output power range of the high-priority charging port is adapted to the fast charging demand, and the output power range of the low-priority charging port is adapted to the conventional charging demand, including: 801. Extracting a spatial density parameter of the fast charging demand in a coverage area of ​​a high-priority charging port according to the spatial distribution of the charging ports with different priorities marked in the priority matching rule; In step 801, the spatial density parameter is a quantitative value of the spatial distribution density of the fast charging demand in the coverage area of ​​the high priority charging port, which is used to characterize the demand intensity.

[0111] In the embodiment of the present application, based on the high-priority port positions marked by the priority matching rules, the kernel density estimation algorithm is used to perform spatial interpolation calculations on the fast charging demand points within the coverage area. The density heat map is generated by the Gaussian kernel function, and the geometric center coordinates and coverage radius of the peak area (density ≥ 1.5 people / ㎡) in the heat map are extracted. The spatial autocorrelation analysis (Moran's I index) is used to verify the spatial aggregation characteristics of the density distribution, remove discrete noise points, and finally output the spatial density parameters including density value, center coordinates and coverage range.

[0112] 802. Based on the spatial density parameter, configure a narrowband resonant frequency range and a small-pitch power transmission coil parameter combination for each high-priority charging port, so that the output power range of the high-priority charging port covers the power range corresponding to the fast charging demand; In step 802, the narrowband resonant frequency range and the small-pitch power transmission coil parameter combination are electromagnetic coupling parameters set to adapt to high-power fast charging requirements, including a specific frequency band resonance range (such as 6.78MHz±0.3MHz) and a short-distance coil layout (≤15cm).

[0113] In the embodiment of the present application, a genetic algorithm (NSGA-II) is used to optimize the multi-objective combination of resonant frequency and coil spacing based on a set of spatial density parameters. The objective function is defined as maximizing output power (≥150kW) and minimizing electromagnetic radiation leakage (≤30dBμV / m), and the constraints include coil temperature rise (≤40°C) and efficiency threshold (≥85%). The parameter fitness surface is constructed through finite element simulation, and the Pareto front optimal solution is selected as the final parameter combination. A dynamic impedance matching network (calibrated based on the Smith chart) is used to adjust the resonant circuit parameters in real time to ensure stable power output.

[0114] 803. Synchronously configure a broadband resonant frequency range and a large-pitch power transmission coil parameter combination for a low-priority charging port, so that the output power range of the low-priority charging port covers the power range corresponding to the conventional charging demand; In step 803, the broadband resonant frequency range and the large-distance power transmission coil parameter combination are electromagnetic coupling parameters set to adapt to conventional charging requirements, including a wide-band resonance range (such as 13.56 MHz ± 2 MHz) and a long-distance coil layout (≥ 30 cm).

[0115] In the embodiment of the present application, in view of the spatial distribution characteristics of the low-priority ports, the ant colony optimization algorithm (ACO) is used to search for a coil layout solution that maximizes the coverage area (≥80㎡) and minimizes energy consumption (≤5kW). The optimal coil spacing is determined by electromagnetic field uniformity analysis (based on the numerical solution of the Poisson equation), and multi-band adaptive switching is achieved in combination with a broadband frequency tuning module (using software-defined radio technology). The output power range of the low-priority port covers the power range corresponding to the conventional charging demand, and a power factor correction circuit (PFC) is introduced to compensate for reactive power loss to ensure that the efficiency under the wide frequency band is ≥75%.

[0116] 804. Collect actual output power fluctuation data of the charging ports with different priorities, and if deviation from the corresponding power interval is detected, recalculate the resonant frequency range and the power transmission coil spacing parameter combination based on the current fluctuation data.

[0117] In step 804, the actual output power fluctuation data is a record of the deviation between the power output value of the wireless charging port during real-time operation and its target range, which is used to trigger parameter re-optimization.

[0118] In the embodiment of the present application, the high-precision power sensor collects the output data of the port, uses the wavelet transform algorithm to decompose the fluctuation signal, and identifies the high-frequency noise and low-frequency trend components. If the trend component continues to deviate from the target interval for more than 3 sampling cycles (15 minutes), the recalculation process of the resonant frequency range and the power transmission coil spacing parameter combination is triggered: based on the Kalman filter algorithm, the demand density change trend is predicted, and the optimization process of steps 802-803 is re-executed in combination with the current spatial density parameters, and the rolling hive control (RHC) strategy is used to dynamically update the parameter combination.

[0119] Here is a specific example: The west side of the underground parking lot of a subway complex is a daytime office fast charging hot zone (7:00-19:00), and the east side is a nighttime residential regular demand zone (20:00-6:00). In step 801, the kernel density estimation shows that the peak value of the fast demand density in the coverage area of ​​the west port P1-P3 is 2.1 people / ㎡, with a coverage radius of 12 meters; in step 802, NSGA-II configures a resonant frequency of 6.78MHz and a coil spacing of 10cm for P1-P3, and the output power is increased to 16 0kW, radiation leakage is controlled at 28dBμV / m; in step 803, ACO optimizes the broadband frequency of 13.56MHz and the spacing of 35cm for the east port P4-P6, covering an area of ​​85㎡, and the single-port power is stabilized at 28kW; in step 804, it is detected at noon that the actual output of P2 fluctuates to 140kW (the target is 160kW, and it continues to deviate from the target range for more than 3 sampling cycles), and the Kalman filter predicts a downward trend in demand density. After re-adjusting the parameters, it recovers to 155kW.

[0120] Steps 801-804 achieve high-precision matching of wireless charging power output and dynamic demand through kernel density quantification, multi-objective parameter optimization, broadband adaptive configuration and fluctuation-driven recalculation. In the scenario of urban mixed-functional areas, the fast charging response speed and conventional equipment utilization rate are significantly improved, electromagnetic interference and energy waste are reduced, and full-link adaptive capabilities are formed from demand perception, parameter optimization to dynamic closed-loop tuning.

[0121] Figure 2 A schematic diagram of a charging pile intelligent site selection and evaluation system is provided for an embodiment of the present application. Figure 2 As shown, the system includes: An acquisition module 21 is used to acquire mobile terminal communication data of a target area, and generate a demand period division model based on the spatiotemporal variation law of the flow density of people in the mobile terminal communication data, wherein the demand period division model is used to distinguish between the fast charging demand corresponding to the daytime period and the conventional charging demand corresponding to the nighttime period; An adjustment module 22 is used to adjust the position of the edge computing device deployed in the target area according to the demand distribution result output by the demand period division model, so that the service range of the edge computing device moves consistently with the peak area of ​​the pedestrian flow density and meets the preset data transmission performance requirements; A control module 23, configured to control the multi-port wireless charging system to switch a charging power output mode based on the charging demand type identified in the position adjustment result, wherein the charging power output mode corresponds to the time periods of the fast charging demand and the regular charging demand; A generation module 24 is used to combine the network layout data after the position adjustment of the edge computing device and the power output mode information of the multi-port wireless charging system to generate a set of candidate charging pile layout locations that simultaneously meet the daytime charging service efficiency and nighttime equipment utilization rate optimization indicators; The screening module 25 is used to verify the underground pipe network safety distance for each location in the candidate charging pile layout location set, and screen out the target charging pile layout plan that meets the preset construction safety standards and matches the demand distribution law of the demand period division model.

[0122] Figure 2 The intelligent site selection and evaluation system for charging piles can be executed Figure 1 The implementation principle and technical effect of the charging pile intelligent site selection and evaluation method described in the embodiment are not repeated here. The specific way in which each module and unit performs operations in the charging pile intelligent site selection and evaluation system in the above embodiment has been described in detail in the embodiment of the method, and will not be elaborated here.

[0123] In one possible design, Figure 2 A charging pile intelligent site selection and evaluation system of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0124] The processing component 32 is used for the above Figure 1 The embodiment provides a charging pile intelligent site selection and evaluation method.

[0125] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0126] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0127] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0128] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0129] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0130] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A charging pile intelligent site selection and evaluation method, characterized in that: include: Acquire mobile terminal communication data of the target area, and generate a demand period division model based on the spatiotemporal variation law of the crowd density in the mobile terminal communication data, wherein the demand period division model is used to distinguish between the fast charging demand corresponding to the daytime period and the conventional charging demand corresponding to the nighttime period; According to the demand distribution result output by the demand period division model, the position of the edge computing device deployed in the target area is adjusted so that the service range of the edge computing device moves in a consistent manner with the peak area of ​​the pedestrian flow density; Based on the charging requirement type identified in the position adjustment result, controlling the multi-port wireless charging system to switch the charging power output mode, the charging power output mode corresponding to the time period of the fast charging requirement and the regular charging requirement; Combining the network layout data after the position adjustment of the edge computing device with the information corresponding to the power output mode of the multi-port wireless charging system, generating a set of candidate charging pile layout locations; A distance verification is performed on each position in the candidate charging pile layout position set, and a target charging pile layout scheme that meets the preset safety standard and matches the demand distribution law of the demand period division model is screened out.

2. The method according to claim 1, characterized in that The combining the network layout data after the position adjustment of the edge computing device and the power output mode information of the multi-port wireless charging system to generate a set of candidate charging pile layout locations includes: extracting the network layout data after the position adjustment of the edge computing device, and identifying a continuous spatial area that meets the daytime charging service efficiency requirements; synchronously acquiring power output mode information of the multi-port wireless charging system, and marking output areas where the power utilization rate during the night time period is continuously higher than the average level; Performing geospatial superposition of the continuous spatial region and the output region, and screening out candidate layout ranges that include both the daytime efficiency region boundary and the nighttime utilization rate core region; Based on the candidate deployment range, the deployable areas are spatially prioritized according to a preset charging pile spacing rule to generate a set of candidate charging pile deployment locations.

3. The method according to claim 1, characterized in that The distance verification is performed on each position in the candidate charging pile layout position set to select a target charging pile layout scheme that meets the preset safety standard and matches the demand distribution law of the demand period division model, including: Obtaining the spatial distance between each location in the candidate charging pile layout location set and the three-dimensional coordinate data of the underground pipe network, and screening out a subset of each location whose straight-line distance to the gas pipeline and the power pipeline gallery is greater than a preset safety distance; Based on the spatial distribution of the fast charging demand hotspots and the conventional charging demand hotspots marked in the demand period division model, the area ratio of the subset covering the two types of demand hotspots is calculated, and the adaptation position that retains the area ratio while meeting the lower limit of the coverage of the daytime fast demand and the nighttime conventional demand is retained; Verify the underground pipe network safety spacing of the adaptation position, eliminate abnormal positions whose spacing with adjacent pipe network positions exceeds the preset spacing interval, and generate a set of intermediate layout plans; A safety assessment is performed based on the historical time period passenger density curve in the intermediate layout plan set, and the test results are similarly matched with the demand distribution law of the demand period division model to screen out the target charging pile layout plan that meets the preset safety standards and matches the demand distribution law.

4. The method according to claim 1, characterized in that: The adjusting the location of the edge computing device deployed in the target area according to the demand distribution result output by the demand time period division model includes: Analyze the demand distribution result output by the demand period division model, extract the high demand area coordinate set in the target area in the current period, and determine the moving trajectory of the peak area of ​​the pedestrian flow density according to the spatiotemporal distribution characteristics of the high demand area coordinate set; Based on the center position of the peak area of ​​each time segment in the moving trajectory, a target deployment position set of the edge computing device is calculated, wherein the straight-line distance between each target position in the target deployment position set and the center of the peak area of ​​the corresponding time segment does not exceed a preset coverage radius; Generate a device migration instruction according to the target deployment location set, drive the edge computing device to migrate to the target location according to the device migration instruction, and establish a communication network with adjacent devices during the migration process, and control the end-to-end transmission delay between nodes in the communication network to be lower than a preset maximum value; An updated coordinate set of the peak area is obtained through the communication network. When it is detected that the coverage overlap rate between the updated coordinate set and the target deployment position set is lower than a preset tolerance threshold, the target deployment position set calculation is re-executed based on the updated coordinate set, and the device migration instruction is synchronously adjusted.

5. The method according to claim 1, characterized in that: The step of controlling the multi-port wireless charging system to switch the charging power output mode based on the charging requirement type identified in the position adjustment result includes: Acquire a charging requirement type marked in the position adjustment result, where the charging requirement type includes a first type of demand signal and a second type of demand signal corresponding to the fast charging requirement and the conventional charging requirement, respectively; According to the spatial distribution density of the charging demand type, a priority matching rule is configured for each charging port of the multi-port wireless charging system, and the priority of the charging port in the area covered by the first type of demand signal is higher than that of the second type of demand signal; Based on the priority matching rule, adjusting the electromagnetic coupling parameter combination of each charging port in the multi-port wireless charging system so that the output power range of the high-priority charging port is adapted to the fast charging requirement, and the output power range of the low-priority charging port is adapted to the conventional charging requirement, wherein the electromagnetic coupling parameter combination includes a resonant frequency matching range and a power transmission coil spacing; The electromagnetic coupling state feedback data of the multi-port wireless charging system is used to periodically detect the degree of compatibility between the actual output power of each charging port and the priority matching rule. If it is detected that the degree of compatibility is lower than the matching rule requirement, the electromagnetic coupling parameter combination is regenerated and the priority matching rule is updated.

6. The method according to claim 1, characterized in that The generating of the demand time period division model based on the spatiotemporal variation law of the crowd flow density in the mobile terminal communication data includes: Extracting the spatial coordinates and timestamps of the terminal devices in the mobile terminal communication data, and constructing a spatiotemporal distribution sequence of the density of human traffic with geographic grids as units; Performing pattern recognition on the fluctuation characteristics of the crowd density in each geographic grid in the crowd density spatiotemporal distribution sequence, extracting a set of grids with short-term high-density aggregation characteristics during the daytime period as candidate areas for fast charging demand, and extracting a set of grids with continuous medium-low density distribution during the nighttime period as candidate areas for conventional charging demand; Based on the spatiotemporal overlap analysis between the candidate areas with fast charging demand and the candidate areas with conventional charging demand, the candidate areas whose spatial coverage areas in adjacent time segments exceed the upper limit of the overlap are merged into demand period division units, and the corresponding charging demand type label is marked for each division unit; According to the label distribution law of the demand time division unit in the historical period, the division granularity of the geographic grid and the determination threshold of the spatiotemporal overlap analysis are modified to generate a demand time division model adapted to the changing characteristics of the flow of people in the target area.

7. The method according to claim 4, characterized in that The generating of the device migration instruction according to the target deployment location set, driving the edge computing device to migrate to the target location according to the device migration instruction, establishing a communication network with adjacent devices during the migration process, and controlling the end-to-end transmission delay between nodes in the communication network to be lower than a preset maximum value, includes: Generate a device migration instruction including a migration path and a migration speed according to a relative distance between each target location in the target deployment location set and a current edge computing device location; Driving the edge computing device to migrate to the target location according to the device migration instruction, detecting the signal strength with the adjacent edge computing device during the migration process, and selecting the adjacent device with the signal strength higher than a preset threshold as a candidate node of the communication network; Based on the location distribution and signal strength data of the candidate nodes, a star-shaped communication network topology centered on the edge computing device is constructed; The end-to-end transmission delay between each topology node in the star communication network topology is monitored. If it is detected that the end-to-end transmission delay exceeds a preset maximum value, the candidate node is reselected and the hop count distribution of the star communication network topology is adjusted.

8. The method according to claim 5, characterized in that The step of adjusting the electromagnetic coupling parameter combination of each charging port in the multi-port wireless charging system so that the output power range of the high-priority charging port adapts to the fast charging requirement and the output power range of the low-priority charging port adapts to the conventional charging requirement includes: Extracting a spatial density parameter of the fast charging demand within a coverage area of ​​a high-priority charging port according to the spatial distribution of the charging ports with different priorities marked in the priority matching rule; Based on the spatial density parameter, a narrowband resonant frequency range and a small-pitch power transmission coil parameter combination are configured for each high-priority charging port, so that the output power range of the high-priority charging port covers the power range corresponding to the fast charging demand; Synchronously configuring a broadband resonant frequency range and a large-pitch power transmission coil parameter combination for a low-priority charging port, so that the output power range of the low-priority charging port covers the power range corresponding to the conventional charging demand; The actual output power fluctuation data of the charging ports with different priorities are collected, and if a deviation from the corresponding power interval is detected, the resonant frequency range and the power transmission coil spacing parameter combination are recalculated based on the current fluctuation data.

9. A charging pile intelligent site selection and evaluation system, characterized in that: include: An acquisition module, used to acquire mobile terminal communication data of a target area, and generate a demand period division model based on the spatiotemporal variation law of the flow density of people in the mobile terminal communication data, wherein the demand period division model is used to distinguish between the fast charging demand corresponding to the daytime period and the conventional charging demand corresponding to the nighttime period; An adjustment module is used to adjust the position of the edge computing device deployed in the target area according to the demand distribution result output by the demand period division model, so that the service range of the edge computing device moves consistently with the peak area of ​​the pedestrian flow density and meets the preset data transmission performance requirements; a control module, configured to control the multi-port wireless charging system to switch a charging power output mode based on the charging demand type identified in the position adjustment result, wherein the charging power output mode corresponds to the time periods of the fast charging demand and the regular charging demand; A generation module, configured to combine the network layout data after the position adjustment of the edge computing device and the power output mode information of the multi-port wireless charging system to generate a set of candidate charging pile layout locations that simultaneously meet the daytime charging service efficiency and nighttime equipment utilization rate optimization indicators; The screening module is used to verify the underground pipe network safety distance for each location in the candidate charging pile layout location set, and screen out the target charging pile layout plan that meets the preset construction safety standards and matches the demand distribution law of the demand period division model.

10. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a charging pile intelligent site selection and evaluation method as described in any one of claims 1 to 8.

Images