Charging station key operation area identification method and device and electronic equipment

Abstract

The invention discloses a charging station key operation area identification method and apparatus, and an electronic device. The method comprises the following steps: constructing a standard grid system of a target city according to a high-precision map of the target city; obtaining multi-source heterogeneous data corresponding to the target city, processing the multi-source heterogeneous data through a large language model to obtain standardized multi-source data, and integrating the standardized multi-source data to a standard grid system to obtain grid standardized data; based on the grid standardized data, performing charging demand prediction simulation on the target city according to a macroscopic and microscopic combined strategy to obtain a charging demand thermodynamic diagram; and acquiring charging facility distribution data and competitive product layout data of the target city, and inputting the charging facility distribution data and the competitive product layout data and the charging demand thermodynamic diagram into the multi-modal large model for evaluation to obtain a charging station key operation area of the target city. According to the technical scheme, accurate identification of the key operation area of the charging station is realized, and site selection scientificity and decision-making efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of charging station site selection technology, and in particular to a method, device and electronic equipment for identifying key operating areas of charging stations. Background Technology

[0002] As a core infrastructure supporting the development of the new energy vehicle industry, charging systems have been included in the national "new infrastructure" key construction scope, and the industry has expanded rapidly under strong policy support. However, affected by multiple factors such as the high uncertainty of electric vehicle user charging behavior and the complexity of market competition, charging operators generally face prominent problems such as supply and demand imbalance and lack of scientific site layout. Traditional charging station planning and decision-making models rely heavily on manual experience and traditional statistical methods, making it difficult to dynamically adapt to changing market demands and complex industry competition. This results in uneven station utilization rates and severely limited return on investment, hindering the sustainable development of the industry.

[0003] To overcome the limitations of traditional models, the industry has gradually applied multidimensional data governance technology and generative artificial intelligence technology. Multidimensional data governance technology, by standardizing and integrating heterogeneous data such as charging demand, geographical location, surrounding prices, and grid load, constructs a comprehensive digital foundation for charging operations, providing data support for decision-making. Generative artificial intelligence, especially Large Language Modeling (LLM), leverages its powerful contextual understanding and knowledge transfer capabilities developed through training on massive amounts of data to provide new paths for analyzing unstructured information (such as planning texts and policy clauses) and complex boundary conditions in planning decisions. However, existing technologies have significant shortcomings: on the one hand, traditional rule-based decision-making models lack deep semantic reasoning capabilities, failing to effectively handle unstructured data and ambiguous supply and demand boundaries, leading to one-sided decision-making basis for identifying high-value areas; on the other hand, the "digital foundation" built by multidimensional data governance and the "deep semantic reasoning capabilities" of large language models have not yet formed effective synergy. The industry lacks intelligent technical means to connect the entire process of "multidimensional data governance - gridded simulation - key area identification - structured recommendations," making the decision-making process unreproducible, difficult to quantify, and lacking interpretability.

[0004] Therefore, how to achieve deep synergy between multidimensional data governance and the intelligent decision-making capabilities of large language models, and build an intelligent method that can effectively handle unstructured data and fuzzy supply and demand boundaries and connect the entire process, so as to provide charging operators with reproducible, quantifiable and scalable key operating area identification and layout decision-making basis, is a key technical problem that the industry urgently needs to solve. Summary of the Invention

[0005] This invention provides a method, device, and electronic equipment for identifying key operating areas of charging stations, so as to achieve accurate identification of key operating areas of charging stations and improve the scientific nature of site selection and decision-making efficiency.

[0006] According to one aspect of the present invention, a method for identifying key operating areas of charging stations is provided, comprising:

[0007] Based on the high-precision map of the target city, a standard grid system for the target city is constructed;

[0008] Obtain multi-source heterogeneous data corresponding to the target city, process the multi-source heterogeneous data through a large language model to obtain standardized multi-source data, and integrate the standardized multi-source data into the standard grid system to obtain grid standardized data;

[0009] Based on the grid-standardized data, a charging demand prediction simulation is performed on the target city according to a strategy combining macro and micro perspectives, resulting in a charging demand heatmap.

[0010] Obtain charging facility distribution data and competitor layout data for the target city, and input the charging facility distribution data, competitor layout data and the charging demand heat map into a multimodal large model for evaluation to obtain the key operating areas of charging stations in the target city.

[0011] According to another aspect of the present invention, a charging station key operating area identification device is provided, comprising:

[0012] The grid system construction module is used to construct a standard grid system for the target city based on a high-precision map of the target city.

[0013] The data processing module is used to acquire multi-source heterogeneous data corresponding to the target city, process the multi-source heterogeneous data through a large language model to obtain standardized multi-source data, and integrate the standardized multi-source data into the standard grid system to obtain grid standardized data.

[0014] The demand simulation module is used to perform charging demand prediction simulation on the target city based on the grid standardized data and according to a strategy that combines macro and micro perspectives, so as to obtain a charging demand heat map.

[0015] The regional assessment module is used to obtain charging facility distribution data and competitor layout data for the target city, and input the charging facility distribution data, competitor layout data and the charging demand heat map into the multimodal large model for assessment to obtain the key operating areas of charging stations in the target city.

[0016] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0017] At least one processor;

[0018] and memory that is communicatively connected to at least one processor;

[0019] The memory stores a computer program that can be executed by at least one processor, which enables the at least one processor to perform the charging station key operating area identification method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the charging station key operating area identification method of any embodiment of the present invention.

[0021] The technical solution of this invention achieves spatial alignment and integration of multi-source heterogeneous data by constructing a unified standard grid system, completes data semantic alignment and intelligent cleaning with the help of a large language model, accurately predicts charging demand through macro-level traffic flow simulation and micro-level user behavior agent coupled simulation, and finally uses a multi-modal large model for multi-dimensional evaluation and identification of key operating areas, forming a fully intelligent technical solution for data governance, demand simulation, area identification, and decision output. This solution solves the technical problem that traditional rule-based decision-making models cannot effectively handle unstructured data and fuzzy supply-demand boundaries, fully exploring the value of complex boundary conditions to improve decision-making basis. On the other hand, it breaks down the disconnect between multi-dimensional data governance and the intelligent decision-making capabilities of large models, making the decision-making process reproducible, quantifiable, and interpretable through fully automated technology, providing charging operators with scientific and accurate layout decision support. It improves the scientificity and accuracy of area identification and demand forecasting, enabling charging station layout decisions to move from traditional experience-driven to intelligent data-driven, significantly optimizing the site selection efficiency and return on investment for charging operators, and providing long-term technical support for their sustainable development.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1A flowchart illustrating a method for identifying key operating areas of charging stations provided in an embodiment of the present invention;

[0025] Figure 2 A flowchart of another method for identifying key operating areas of charging stations provided in an embodiment of the present invention.

[0026] Figure 3 This invention provides a four-layer architecture design diagram for intelligent decision-making in charging station site selection;

[0027] Figure 4 A flowchart illustrating another method for identifying key operating areas of charging stations provided in this embodiment of the invention;

[0028] Figure 5 A flowchart of gridded simulation of charging demand based on traffic-side data provided in an embodiment of the present invention;

[0029] Figure 6 A flowchart for designing large-scale prompt words based on thought chains is provided for embodiments of the present invention;

[0030] Figure 7 This is a schematic diagram of the structure of a key operating area identification device for charging stations provided in an embodiment of the present invention;

[0031] Figure 8 A schematic diagram of the structure of an electronic device for implementing the charging station key operating area identification method of this embodiment of the invention. Detailed Implementation

[0032] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0033] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0034] It should be noted that the collection, gathering, updating, analysis, processing, use, transmission, and storage of user personal information involved in this disclosed technical solution all comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken to prevent unauthorized access to user personal information data and to safeguard user personal information security, network security, and national security.

[0035] Figure 1 This is a flowchart illustrating a method for identifying key operating areas of charging stations according to an embodiment of the present invention. This embodiment is applicable to the site selection of charging stations. The method can be executed by a device for identifying key operating areas of charging stations, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method specifically includes the following steps:

[0036] S110. Based on the high-precision map of the target city, construct a standard grid system for the target city.

[0037] The target cities can be understood as those cities where key operating areas for charging stations need to be identified.

[0038] In some embodiments, a standard grid system for the target city is constructed based on a high-precision map of the target city, including: determining a preset grid division scale; dividing the target city into two or more urban grids based on the high-precision map of the target city and the preset grid division scale; and forming a standard grid system based on the two or more urban grids.

[0039] Each city grid is associated with a unique geocode and the latitude and longitude coordinates of its center point.

[0040] Specifically, a unified spatial grid system can be constructed as the foundation for data integration and analysis. First, a standard grid system covering the entire target city is established using a high-precision urban map. In this embodiment, the grid scale can be flexibly set in advance, dividing the spatial grid into sizes ranging from 200m×200m to 1km×1km based on the density of the urban built-up area. Each grid is assigned a unique geographic code and the latitude and longitude coordinates of its center point. This grid system serves as a unified analysis unit, used to subsequently carry various multi-source information such as traffic flow, POI distribution, and charging heatmaps, ensuring that all heterogeneous data can be accurately located and aligned spatially.

[0041] It should also be noted that spatial partitioning units are not limited to square grids. Hexagonal grids (as they are generally superior to rectangles in path connectivity analysis), irregular polygons (such as traffic zones (TAZs) formed based on urban road networks), or geofencing units based on administrative divisions (street / community level) can also be used. Alternatively, instead of using fixed-size grids, multi-level adaptive grids can be used based on the level of activity or data density of the area (e.g., a 50m fine grid for the core area and a 2km coarse grid for the suburbs), or a spatial index structure based on quadtrees can be employed.

[0042] S120. Obtain multi-source heterogeneous data corresponding to the target city, process the multi-source heterogeneous data through a large language model to obtain standardized multi-source data, and integrate the standardized multi-source data into a standard grid system to obtain grid standardized data.

[0043] In some embodiments, multi-source heterogeneous data corresponding to the target city is obtained, and the multi-source heterogeneous data is processed by a large language model to obtain standardized multi-source data. This includes: inputting the multi-source heterogeneous data into the large language model and constructing data governance prompt words; using the data governance prompt words to guide the large language model to perform semantic alignment and intelligent cleaning operations on the multi-source heterogeneous data to obtain standardized multi-source data.

[0044] The multi-source heterogeneous data includes transportation data, facility data, and energy trading data. Transportation data includes urban road network structure and traffic flow monitoring data for key road sections. Facility data includes point of interest (POI) data (such as shopping malls, hospitals, and office parks), distribution of existing charging facilities, and their utilization rate. Energy and trading data includes regional power grid capacity constraint information and historical charging order data.

[0045] In this embodiment, the specific operations performed on multi-source heterogeneous data include semantic alignment (e.g., unifying "parking garage" and "underground parking garage" into the semantic meaning of "parking facility") and intelligent cleaning (identifying and removing abnormal noisy data). Taking the governance of parking lot point of interest (POI) data as an example, the input of the large model is multi-source heterogeneous raw unstructured data entries, such as descriptions including "parking garage on the B2 floor of a shopping mall", "roadside temporary parking space", and non-physical facilities such as "system test point 123". At this point, the constructed prompt is: "You are an urban data governance expert. Please map the following input data to a standard category (such as 'public parking lot', 'street parking space') based on contextual semantics, extract key attributes (such as 'indoor', 'outdoor'), and identify and remove entries containing semantics such as 'test', 'discontinued', and 'demolished' as noise. The final output format is standardized JSON data containing 'standard name', 'facility type', and 'validity identifier'." Based on this prompt, the final output of the large model accurately aligns "B2 level parking garage" to the standard fields "facility type: public parking lot" and "attribute: indoor," while marking "system test point" as "invalid" and automatically filtering it. Through this step, a standardized fusion data base is constructed, namely standardized multi-source data. Then, the standardized multi-source data is integrated into a standard grid system to obtain grid standardized data. Preferably, the standardized multi-source data can be associated with the urban grid.

[0046] S130. Based on grid-standardized data, a charging demand prediction simulation is performed on the target city according to a strategy combining macro and micro perspectives to obtain a charging demand heat map.

[0047] In this step, charging demand prediction simulations can be performed based on a large model, that is, using grid-normalized data to generate a charging demand heatmap through a combination of macro and micro methods:

[0048] At the macro level: a macro traffic simulation model can be introduced, which can construct an origin-destination (OD) transfer probability matrix based on urban functional zoning to simulate the macro flow trend of traffic between urban grids;

[0049] At the micro level: Deploy micro-level user agents driven by a large language model. Each agent simulates an electric vehicle driver, making human-like behavioral decisions based on the vehicle's remaining battery charge (SOC), current location, trip purpose, and personal preferences (such as price sensitivity and fast charging preference). The system couples macro-level traffic flow with the charging decisions of micro-level agents to deduce the total charging demand for each grid within a specific future time period, thereby generating a grid-level charging demand heatmap.

[0050] S140. Obtain charging facility distribution data and competitor layout data for the target city. Input the charging facility distribution data, competitor layout data, and charging demand heat map into the multimodal large model for evaluation to obtain the key operating areas of charging stations in the target city.

[0051] In some embodiments, charging facility distribution data, competitor layout data, and charging demand heatmaps are input into a multimodal large model for evaluation to obtain the key operating areas of charging stations in the target city. This includes: inputting charging facility distribution data, competitor layout data, and charging demand heatmaps into a multimodal large model for evaluation; setting the multimodal large model as a charging network planning expert to guide the multimodal large model to score each city grid according to preset dimensions; and determining the key operating areas of charging stations in the target city based on the scoring results of each city grid.

[0052] Understandably, after obtaining the charging demand heatmap, further grid-level construction potential assessment and area identification can be carried out. Specifically, the generated charging demand heatmap, combined with existing charging facility distribution and competitive landscape data, can be input into a multimodal large model for comprehensive evaluation.

[0053] This step can employ role-based prompting technology, setting the multimodal large model as a "senior charging network planning expert." The model utilizes spatial semantic reasoning capabilities to score each grid in multiple dimensions (including demand density, competition intensity, and charging convenience), identifying key areas with high overall scores. This process represents a leap from simple data display to intelligent decision-making.

[0054] In some optional embodiments, this embodiment can output structured layout suggestions and verify the closed loop. For the identified key operating areas, the large model ultimately outputs structured layout suggestions, including:

[0055] Recommended coordinates: specific latitude and longitude coordinates or recommended plots of land;

[0056] Charging station configuration: Recommended fast / slow charging ratio and power rating;

[0057] Natural Language Reasoning: Generate an interpretable site selection report (e.g., "This grid is located at the junction of a core business district and a residential area, with high demand for energy replenishment during the evening rush hour, and no large competing stations within 1km"). Figure 2 The diagram shown is a flowchart of another method for identifying key operating areas of charging stations provided by an embodiment of the present invention.

[0058] The technical solution of this invention, by constructing a gridded spatial data base and a simulation prediction framework based on a large model, achieves full-process automation from multi-dimensional data governance to intelligent identification of key operating areas. Its beneficial effects are that, by leveraging the deep reasoning of complex spatial semantics, accurate simulation of macro and micro behaviors, and collaboration with dynamic data through the large model, the scientificity and accuracy of area identification and demand prediction are improved. This enables charging station layout decisions to move from traditional experience-driven to intelligent data-driven, significantly optimizing the site selection efficiency and return on investment for charging operators, and providing long-term technical support for their sustainable development.

[0059] Based on the above embodiments, this invention provides a four-layer architecture design for intelligent decision-making in charging station site selection. This architecture can realize the key operating area identification method for charging stations described in the aforementioned embodiments, and its design principle is as follows: Figure 3 As shown:

[0060] First layer: Multi-source spatial cognition and data input layer.

[0061] This layer is responsible for the comprehensive characterization and standardized input of the urban operating environment. The system collects data such as spatial geographic information, socio-economic characteristics, current infrastructure status, historical charging heat maps and charging demand forecasts around a unified spatial grid system. It uniformly encodes elements such as latitude and longitude, land use type, population and vehicle scale, road grade, parking facilities, location and utilization rate of existing charging stations, and grid access capacity for each grid.

[0062] Beyond structured data, this layer also parses unstructured materials such as planning texts, policy clauses, and investment promotion schemes, extracting semantic tags such as "construction restrictions," "encouraged directions," and "special sensitive areas," and binding them to corresponding grids. Missing data filling, anomaly removal, coordinate correction, and time alignment are performed on the aforementioned multi-source data. A standardized data view is constructed according to "grid ID × time granularity," enabling the upper layer to directly access spatial, temporal, and semantic information within the same coordinate system.

[0063] The second layer: a large-scale model reasoning construction layer based on thought chains.

[0064] This layer is designed with "how to think" in mind, responsible for building the reasoning path and intermediate conclusion structure of the large model. The system assigns roles such as "charging network planning expert" to the large model in the prompts, providing clear task objectives, preferred planning principles, and key indicators to focus on, guiding the model to examine each grid from an expert's perspective.

[0065] The thought process revolves around "spatial location - traffic and demand - competitors and constraints - comprehensive judgment": the model is required to first determine the region type based on geographical location and functional zone attributes, then analyze the energy replenishment pressure by combining traffic flow intensity, historical orders and demand forecast results, then examine the existing charging station layout, power access conditions and policy restrictions, and finally give a preliminary opinion on whether the grid is qualified to enter the candidate set.

[0066] To ensure traceability, the prompts also incorporate evidence citation and collaborative verification mechanisms: when the model provides each intermediate conclusion, it must indicate the data field or table row number on which it is based; for links involving hard constraints such as power grid capacity and construction red lines, a small model or rule engine can be called for quick verification, and the large model completes comprehensive reasoning based on this.

[0067] The third layer: Explainable scoring and business priority output layer.

[0068] This layer transforms the candidate grid formed in the second layer into quantitative results that can be directly used for business decisions. The system inputs indicators such as demand intensity, transportation accessibility, construction difficulty, competitive pressure, electricity price level, grid connection conditions, and expected investment returns, along with the weight configuration given by the operator, into the scoring module and the large model to generate quantitative scores for each indicator and overall confidence levels.

[0069] Based on quantification, the system calculates a comprehensive score for each grid and categorizes it into three levels: "Key Deployment," "Regular Deployment," and "Delayed Deployment," providing a ranking of operational priorities. Simultaneously, the large model generates concise textual descriptions following a thought process chain, highlighting the main driving factors for high scores (such as "high demand, few competitors, and convenient power access") and the key shortcomings that limit scores, forming a structured and easily comparable decision-making basis.

[0070] The output of this layer is presented in two forms: first, grid-level indicators and comprehensive score tables, which can be used for batch statistics and filtering; second, interpretable descriptions for individual regions, which can be directly embedded in feasibility study reports, project proposals, and internal review materials.

[0071] The fourth layer: the iterative layer of manual review and knowledge feedback.

[0072] This layer enables closed-loop interaction between planners and the large model, allowing for human-machine collaborative correction of the aforementioned reasoning and scoring results. Planning and operations personnel review each candidate grid on the system interface, providing "adopt," "postpone," or "reject" opinions for each recommendation based on factors such as local planning boundaries, project reserves, and construction feasibility. A brief reason is recorded when rejecting or adjusting, such as "difficulty in grid connection," "currently under construction," or "land attributes do not currently support charging purposes."

[0073] The system compiles these manually labeled information, along with the original grid features, model output scores, and written reasons, into feedback samples. These samples are then summarized by the larger model to identify common patterns in adopted projects (such as typical locations, load ranges, and competitive landscapes) and high-frequency constraints and risks in rejected projects, culminating in "rules of experience" and "warning examples." Based on this, the system automatically adjusts the focus, scoring weights, and constraint settings in the thought chain prompts, generating new rule bases and prompt templates.

[0074] The updated rules and templates are applied again to the second and third layers, enabling a backflow from the feedback layer to the inference and output layers. Figure 3 The feedback arrows are represented from bottom to top. As multiple iterations unfold, the large model's understanding of the preferences of the city or operator deepens, and the recommendation results gradually converge with the actual implementation.

[0075] Through the coordinated operation of the above four-layer architecture, this invention organically integrates multi-source data governance, complex scenario understanding, interpretable quantitative scoring, and human-machine collaborative feedback. It can not only finely depict the distribution of charging demand and location value at the city scale, but also continuously correct the reasoning logic and parameter configuration of the large model in continuous practice. Ultimately, it provides operators with key operating areas and site layout suggestions that can be directly implemented, thereby significantly improving the scientificity, accuracy, and interpretability of charging station site selection.

[0076] Figure 4 This is a flowchart illustrating another method for identifying key operating areas of charging stations provided by an embodiment of the present invention. Based on the above embodiments, this embodiment further optimizes the process of performing charging demand prediction simulations on target cities using a strategy combining macro and micro perspectives to obtain a charging demand heatmap. For example... Figure 4 As shown, the method specifically includes the following steps:

[0077] S410. Based on the high-precision map of the target city, construct a standard grid system for the target city.

[0078] S420. Obtain multi-source heterogeneous data corresponding to the target city, process the multi-source heterogeneous data through a large language model to obtain standardized multi-source data, and integrate the standardized multi-source data into a standard grid system to obtain grid standardized data.

[0079] S430. Based on grid-standardized data, construct the origin-end point transfer probability matrix corresponding to the target city, and simulate the macro-flow trend of traffic flow in the standard grid system based on the origin-end point transfer probability matrix to obtain macro-traffic data.

[0080] In some embodiments, based on grid-standardized data, a start-end point transfer probability matrix corresponding to the target city is constructed, and the macro-flow trend of traffic flow in the standard grid system is simulated based on the start-end point transfer probability matrix to obtain macro-traffic data. This includes: determining the urban functional zoning and traffic survey statistics of the target city based on grid-standardized data, and constructing the start-end point transfer probability matrix based on the urban functional zoning and traffic survey statistics; and simulating the traffic flow migration trajectory of the urban grid in each time period based on the start-end point transfer probability matrix to obtain macro-traffic data.

[0081] like Figure 5 The diagram shown is a flowchart of a gridded simulation of charging demand based on traffic-side data, provided in an embodiment of the present invention. An origin-destination (OD) transition probability matrix can be constructed based on urban functional zoning and traffic survey statistics to simulate vehicle flow patterns within the city, forming the underlying simulation framework for traffic flow. This model captures the travel patterns of different functional areas (such as residential areas to office areas) at different times, providing basic traffic flow input for micro-simulation. Matrix elements express Vehicles move from the grid at specific times (divided into 1-hour time units). Transfer to grid The probability is calculated as follows:

[0082] Formula 1: OD transition probability matrix

[0083]

[0084] in, Indicates in Time Grid To grid Actual monitored traffic flow This represents the total number of grid cells.

[0085] S440. Based on grid-standardized data, deploy micro-user agents to simulate the charging decision-making behavior of electric vehicle drivers and obtain charging decision results.

[0086] In some embodiments, based on grid-standardized data, micro-user agents are deployed to simulate the charging decision-making behavior of electric vehicle drivers and obtain charging decision results. This includes: assigning basic attributes to each micro-user agent based on grid-standardized data; embedding charging decision rules into a large language model; inputting the basic attributes into the large language model; and enabling the large model to infer the charging decision result corresponding to each micro-user agent based on the charging decision rules.

[0087] The basic attributes include at least one of the following: vehicle type, battery capacity, real-time remaining battery power, current grid location, destination, price sensitivity, and charging preference.

[0088] S450, based on macro traffic flow data and charging decision results, deduces the total charging demand of each city grid in the future time period, and generates a charging demand heat map corresponding to the target city based on the total charging demand.

[0089] In this embodiment of the invention, a micro-level user behavior agent can be established. That is, an intelligent agent driven by a large language model is introduced as a micro-level decision-making unit. Each agent makes a human-like charging decision based on its vehicle's real-time battery status, the spatiotemporal context, personal preferences, and information about surrounding facilities. The large language model acts as the "brain" of the intelligent agent, utilizing its contextual understanding capabilities to simulate the complex behavior of users facing fuzzy decisions.

[0090] 1. Agent initialization: Assign basic attributes to each Agent, including core parameters such as vehicle type, battery capacity, real-time remaining battery power, current grid location, destination, price sensitivity, and charging preference;

[0091] 2. Decision Logic Embedding: Through prompt word engineering, the charging decision rules are transformed into executable instructions for the model. Example prompt word template: "You are an electric vehicle user, currently located at grid X, with 25% battery remaining, and your destination is grid Y (30km away). There are two charging stations within 1km (Station A: fast charging 1.8 yuan / kWh, Station B: slow charging 0.8 yuan / kWh). You are price-sensitive. Do you need to charge? If so, which station should you choose?"

[0092] 3. Dynamic decision output: The Agent outputs charging decisions, charging station selection, and estimated charging time based on real-time scenario parameters (such as triggering forced charging when the remaining battery is below 20% or replenishing energy in advance if there are no charging facilities in the destination grid) through semantic reasoning of a large model.

[0093] Furthermore, coupled simulation and requirement mapping can be performed, specifically:

[0094] 1. Dynamic Coupling Mechanism: The inter-grid vehicle flow migration trajectory generated by the macro-level traffic flow model is dynamically correlated with the real-time location and power status of the micro-level Agent. As the Agent moves between grids with the macro-level traffic flow, it consumes power in real time according to the vehicle energy consumption model;

[0095] 2. Charging demand trigger: When the Agent meets the charging triggering conditions, it triggers a charging demand event and records information such as the occurrence grid, the required amount of electricity, and the expected charging period.

[0096] 3. Precise grid mapping: Through spatial indexing technology, each charging demand event is precisely associated with the geospatial grid in which it occurs, realizing the precise conversion from "macro traffic flow" to "grid-level charging demand".

[0097] Finally, a gridded demand distribution can be generated, specifically:

[0098] Within the set simulation period, all charging demand events generated by all agents are aggregated, statistically analyzed and quantified by grid, and the total charging demand for each grid is calculated. Ultimately, a visualized spatial distribution map (heat map) of urban charging demand reflecting spatial differences is generated, providing direct data input for subsequent area identification. The calculation formula is as follows:

[0099] Formula 2: Calculation of total grid charging demand

[0100]

[0101] in, for The charging power required by the grid at any given time. for Time is located in the grid The set of Agents To provide Agenta with the necessary power for charging, This is a function to indicate the charging event (1 indicates that charging has been triggered, and 0 indicates that charging has not been triggered).

[0102] The grid's cumulative charging load is obtained by summing up the grid's power demand for each time period within the simulation cycle. Calculate as follows:

[0103] Formula 3: Total Grid Load Aggregation

[0104]

[0105] in, Indicates the entire simulation cycle Inner grid The cumulative charging load generated is a key indicator for assessing the potential construction value of the grid. The final result is a visualized charging demand heatmap, organized by grid and reflecting spatial differences, providing direct data input for subsequent identification of key areas.

[0106] S460: Obtain charging facility distribution data and competitor layout data for the target city. Input the charging facility distribution data, competitor layout data, and charging demand heat map into the multimodal large model for evaluation to obtain the key operating areas of charging stations in the target city.

[0107] In an optional embodiment, the present invention also provides a large-scale prompt word design based on thought chains, as shown in the appendix. Figure 6 As shown, it includes: task objective chain, geographic semantic chain, evidence citation chain, reflection and error correction chain, and output format chain.

[0108] S1: Construct the task target chain

[0109] The prompt clearly states the rules: "You are now a site selection expert. Your task is not to give me results directly, but to strictly follow the order of 'data → features → indicators → decisions'. You must first read the raw data I give you (such as pedestrian and vehicle flow values), convert them into scoring indicators, and finally sort them according to the scores."

[0110] Specifically, the prompt should include the instruction: "No direct assumptions allowed. You must demonstrate how you calculated the final priority from the numbers in the original table."

[0111] S2: Establish geographic semantic chains

[0112] Leveraging the common-sense nature of large-scale models, prompts guide them to translate geographical information into electricity demand. The prompts pre-define the reasoning logic: "Please analyze in the following order: first, determine the functional area (e.g., commercial area); second, assess transportation convenience; and finally, consider the number of surrounding shops to infer the charging demand here."

[0113] Specifically, the model is told, for example: "If you see data that says 'core business district' and 'near the subway,' you must mark it as 'high charging potential'; if you see 'wasteland' and 'no shops,' you must mark it as 'low potential.'"

[0114] S3: Introducing a chain of evidence citation

[0115] To prevent large models from fabricating data, verification requirements are added to the prompts. The prompts explicitly state: "For each of your conclusions, you must find the corresponding line number or paragraph in the original file. The format requirement is: Conclusion + [Basis: Table xx, line xx]".

[0116] Specifically, the model is forced to operate in a "read-as-you-go" mode. For example, when the model analyzes that "parking is difficult in area A," it must immediately follow it with "[Based on: row 12 of the parking data table]". If it cannot find this row in the table, the prompt word will prevent it from writing this sentence.

[0117] S4: Implement the reflection and correction chain

[0118] Before the model outputs its final ranking, force it to stop and check for logical inconsistencies. Add a "reflection instruction" at the end of the prompt: "Before outputting the results, check yourself: Are there any anomalies where 'predicted battery levels are high' but in reality, 'there are very few shops' in that area? If so, your analysis is wrong, please correct it immediately."

[0119] Specifically, an "alarm" was set up for the large model. Once it finds that its reasoning results conflict with common sense (such as predicting high power in an uninhabited area), it must automatically reduce the score of that area and eliminate the error.

[0120] S5: Standard Output Format Chain

[0121] To ensure accuracy, strict formatting guidelines are in place for the output. The prompt should include the following: "Your answer must be divided into two parts. First, write out your thought process and reasoning from S1 to S4; second, list the 'TOP-N' areas with the highest scores."

[0122] Specifically, it outputs both the ranking and the complete reasoning chain. If there's a problem with the ranking, go directly to the "Thinking Process" section in Part 1 to find the incorrect step.

[0123] The technical solution of this invention, by constructing a gridded spatial data base and a simulation prediction framework based on a large model, achieves full-process automation from multi-dimensional data governance to intelligent identification of key operating areas. Its beneficial effects are that, by leveraging the deep reasoning of complex spatial semantics, accurate simulation of macro and micro behaviors, and collaboration with dynamic data through the large model, the scientificity and accuracy of area identification and demand prediction are improved. This enables charging station layout decisions to move from traditional experience-driven to intelligent data-driven, significantly optimizing the site selection efficiency and return on investment for charging operators, and providing long-term technical support for their sustainable development.

[0124] Figure 7 This is a schematic diagram of a key operating area identification device for charging stations provided in an embodiment of the present invention. Figure 7 As shown, the device includes:

[0125] The grid system construction module 710 is used to construct a standard grid system for the target city based on a high-precision map of the target city.

[0126] The data processing module 720 is used to acquire multi-source heterogeneous data corresponding to the target city, process the multi-source heterogeneous data through a large language model to obtain standardized multi-source data, and integrate the standardized multi-source data into the standard grid system to obtain grid standardized data.

[0127] The demand simulation module 730 is used to perform charging demand prediction simulation on the target city based on the grid standardized data and according to a strategy that combines macro and micro perspectives, so as to obtain a charging demand heat map.

[0128] The regional assessment module 740 is used to obtain the charging facility distribution data and competitor layout data of the target city, and input the charging facility distribution data, competitor layout data and the charging demand heat map into the multimodal large model for evaluation to obtain the key operating areas of the charging stations in the target city.

[0129] In some optional implementations, the mesh architecture construction module 310 includes:

[0130] A grid division unit is used to determine a preset grid division scale. Based on the high-precision map of the target city and the preset grid division scale, the target city is divided into two or more urban grids, wherein each urban grid is associated with a unique geographic code and the latitude and longitude coordinates of its center point.

[0131] System components are used to form the standard grid system based on two or more of the aforementioned urban grids.

[0132] In some alternative implementations, the data processing module 720 includes:

[0133] A data input unit is used to input the multi-source heterogeneous data into the large language model and construct data governance prompt words, wherein the multi-source heterogeneous data includes transportation data, facility data and energy trading data;

[0134] The standardization processing unit is used to guide the large language model to perform semantic alignment and intelligent cleaning operations on the multi-source heterogeneous data through the data governance prompt words, so as to obtain the standardized multi-source data.

[0135] In some alternative implementations, the requirement simulation module 730 includes:

[0136] The macro-traffic flow simulation unit is used to construct the origin-end point transfer probability matrix corresponding to the target city based on the grid-standardized data, and to simulate the macro-traffic flow trend of traffic in the standard grid system based on the origin-end point transfer probability matrix to obtain macro-traffic flow data.

[0137] The micro-decision simulation unit is used to deploy a micro-user agent based on the grid-standardized data, and simulate the charging decision behavior of electric vehicle drivers through the micro-user agent to obtain charging decision results;

[0138] The demand heat map generation unit is used to deduce the total charging demand of each city grid in the future time period based on the macro traffic flow data and the charging decision results, and generate a charging demand heat map corresponding to the target city based on the total charging demand.

[0139] In some alternative implementations, the macroscopic traffic flow simulation unit includes:

[0140] The matrix construction sub-unit is used to determine the urban functional zoning and traffic survey statistics of the target city based on the grid-standardized data, and to construct the origin-end point transition probability matrix based on the urban functional zoning and traffic survey statistics.

[0141] The traffic flow trajectory simulation subunit is used to simulate the traffic flow migration trajectory of the urban grid in each time period based on the origin-end transition probability matrix, so as to obtain the macro traffic flow data.

[0142] In some optional implementations, the micro-decision simulation unit includes:

[0143] An attribute allocation subunit is used to allocate basic attributes to each micro-user agent based on the grid standardized data, wherein the basic attributes include at least one of vehicle type, battery capacity, real-time remaining power, current grid location, destination, price sensitivity, and charging preference;

[0144] The decision reasoning subunit is used to embed the charging decision rules into the large language model, input the basic attributes into the large language model, and enable the large language model to infer the charging decision result corresponding to each micro-user agent based on the charging decision rules.

[0145] In some alternative implementations, the demand heat generation unit includes:

[0146] The total demand calculation subunit is used to determine the required charging amount corresponding to each micro user agent based on the charging decision result, and to determine the total charging demand of each city grid in the future time period based on the macro traffic flow data and the required charging amount.

[0147] A heatmap generation subunit is used to generate the charging demand heatmap based on the total charging demand of each of the city grids.

[0148] In some alternative implementations, the area assessment module 740 includes:

[0149] The data input evaluation unit is used to input the charging facility distribution data, competitor layout data and the charging demand heat map into the multimodal large model for evaluation;

[0150] The grid scoring unit is used to set the multimodal large model as a charging network planning expert role and guide the multimodal large model to score each of the city grids according to preset dimensions.

[0151] The key area determination unit is used to determine the key operating areas of charging stations in the target city based on the scoring results of each city grid.

[0152] The charging station key operating area identification device provided in the embodiments of the present invention can execute the charging station key operating area identification method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method.

[0153] Figure 8 This is a schematic diagram of the structure of an electronic device for implementing the charging station key operating area identification method according to embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0154] like Figure 8 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded into the RAM 13 from storage unit 18. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0155] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0156] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the charging station key operating area identification method.

[0157] In some embodiments, the charging station key operating area identification method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the charging station key operating area identification method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the charging station key operating area identification method by any other suitable means (e.g., by means of firmware).

[0158] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0159] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0160] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0161] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0162] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0163] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0164] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0165] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A charging station key business area identification method, characterized by, include: Based on the high-precision map of the target city, a standard grid system for the target city is constructed; Obtain multi-source heterogeneous data corresponding to the target city, process the multi-source heterogeneous data through a large language model to obtain standardized multi-source data, and integrate the standardized multi-source data into the standard grid system to obtain grid standardized data; Based on the grid-standardized data, a charging demand prediction simulation is performed on the target city according to a strategy combining macro and micro perspectives, resulting in a charging demand heatmap. Obtain charging facility distribution data and competitor layout data for the target city, and input the charging facility distribution data, competitor layout data and the charging demand heat map into a multimodal large model for evaluation to obtain the key operating areas of charging stations in the target city.

2. The method of claim 1, wherein, The step of constructing a standard grid system for the target city based on a high-precision map of the target city includes: A preset grid division scale is determined, and based on the high-precision map of the target city and the preset grid division scale, the target city is divided into two or more urban grids, wherein each urban grid is associated with a unique geographic code and the latitude and longitude coordinates of its center point; The standard grid system is composed of two or more of the aforementioned urban grids.

3. The method according to claim 1, characterized in that, The process of acquiring multi-source heterogeneous data corresponding to the target city, and processing the multi-source heterogeneous data using a large language model to obtain standardized multi-source data includes: The multi-source heterogeneous data is input into the large language model, and data governance prompt words are constructed. The multi-source heterogeneous data includes transportation data, facility data, and energy trading data. The data governance prompts guide the large language model to perform semantic alignment and intelligent cleaning operations on the multi-source heterogeneous data to obtain the standardized multi-source data.

4. The method according to claim 2, characterized in that, Based on the grid-standardized data, a charging demand prediction simulation is performed on the target city using a strategy combining macro and micro perspectives to obtain a charging demand heatmap, including: Based on the grid-standardized data, a start-end point transfer probability matrix corresponding to the target city is constructed, and the macro-flow trend of traffic flow in the standard grid system is simulated based on the start-end point transfer probability matrix to obtain macro-traffic data; Based on the grid-standardized data, a micro-user agent is deployed, and the charging decision behavior of electric vehicle drivers is simulated through the micro-user agent to obtain the charging decision results; Based on the macro traffic flow data and the charging decision results, the total charging demand of each city grid in the future time period is deduced, and a charging demand heat map corresponding to the target city is generated based on the total charging demand.

5. The method according to claim 4, characterized in that, Based on the standardized grid data, a origin-destination transfer probability matrix corresponding to the target city is constructed. Then, based on the origin-destination transfer probability matrix, the macroscopic flow trend of traffic flow within the standardized grid system is simulated to obtain macroscopic traffic flow data, including: Based on the grid-standardized data, the urban functional zoning and traffic survey statistics of the target city are determined, and the origin-end point transition probability matrix is ​​constructed based on the urban functional zoning and traffic survey statistics. Based on the origin-end point transition probability matrix, the traffic flow migration trajectory of the urban grid in each time period is simulated to obtain the macro traffic flow data.

6. The method according to claim 4, characterized in that, Based on the grid-standardized data, a micro-user agent is deployed to simulate the charging decision-making behavior of electric vehicle drivers, and the charging decision results are obtained, including: Based on the grid-standardized data, basic attributes are assigned to each micro-user agent, wherein the basic attributes include at least one of vehicle type, battery capacity, real-time remaining battery power, current grid location, destination, price sensitivity, and charging preference; The charging decision rules are embedded into the large language model, and the basic attributes are input into the large language model, so that the large model can infer the charging decision result corresponding to each micro user agent based on the charging decision rules.

7. The method according to claim 4, characterized in that, The process of deducing the total charging demand for each of the city grids in the future time period based on the macro-level traffic flow data and the charging decision results, and generating a charging demand heatmap for the target city based on the total charging demand, includes: Based on the charging decision results, the required charging amount for each micro-user agent is determined. Based on the macro-traffic data and the required charging amount, the total charging demand for each city grid in the future time period is determined. The charging demand heatmap is generated based on the total charging demand of each of the aforementioned city grids.

8. The method according to claim 1, characterized in that, The process involves inputting the charging facility distribution data, competitor layout data, and the charging demand heatmap into a multimodal large model for evaluation to determine the key operating areas for charging stations in the target city, including: The charging facility distribution data, competitor layout data, and charging demand heatmap are input into the multimodal large model for evaluation; The multimodal large model is set as a charging network planning expert role, guiding the multimodal large model to score each of the city grids according to preset dimensions; Based on the scoring results of each city grid, the key operating areas for charging stations in the target city are determined.

9. A device for identifying key operating areas of charging stations, characterized in that, include: The grid system construction module is used to construct a standard grid system for the target city based on a high-precision map of the target city. The data processing module is used to acquire multi-source heterogeneous data corresponding to the target city, process the multi-source heterogeneous data through a large language model to obtain standardized multi-source data, and integrate the standardized multi-source data into the standard grid system to obtain grid standardized data. The demand simulation module is used to perform charging demand prediction simulation on the target city based on the grid standardized data and according to a strategy that combines macro and micro perspectives, so as to obtain a charging demand heat map. The regional assessment module is used to obtain charging facility distribution data and competitor layout data for the target city, and input the charging facility distribution data, competitor layout data and the charging demand heat map into the multimodal large model for assessment to obtain the key operating areas of charging stations in the target city.

10. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, which is then executed by the at least one processor to enable the at least one processor to perform the charging station key operating area identification method according to any one of claims 1-8.

Images