Intelligent Site Selection Base Station Management System and Method

Abstract

This invention relates to the field of intelligent site selection technology for new energy vehicle charging facilities, and discloses a smart site selection base station management system and method. The system includes: a multi-dimensional urban data acquisition and preprocessing module, a knowledge base construction module, an urban evolution dynamics simulation module, a future scenario generation module, a charging network optimization model construction module, and a robust optimization solution module. The method includes the following steps: collecting and constructing a knowledge base of multi-dimensional urban data; then using multi-agent modeling and spatiotemporal graph neural networks to simulate urban dynamics to generate multiple future scenario sets; finally, based on these scenarios, constructing a multi-objective robust optimization model and solving to generate a set of Pareto-optimal robust charging network planning schemes. This invention improves the overall scientific rigor, foresight, and robustness of charging base station site selection planning, thereby effectively reducing its investment decision-making risks.

Description

Technical Field

[0001] This invention relates to the field of intelligent site selection technology for new energy vehicle charging facilities, and in particular to a smart site selection base station management system and method. Background Technology

[0002] With the global energy structure transformation and the advancement of "dual-carbon" goals, and the rapid development of the new energy vehicle industry, the planning and construction of charging networks, as a key infrastructure, are of paramount importance. The scientific site selection of charging base stations directly impacts user experience and return on investment; therefore, how to conduct precise and forward-looking intelligent site selection for charging base stations has become an important issue for the industry's development.

[0003] In existing technologies, the selection of charging base station sites largely relies on static data from Geographic Information Systems (GIS) combined with expert experience for decision-making. This method uses overlay layers such as Points of Interest (POIs) and road networks for visual analysis to filter candidate sites, and is a commonly used technical approach currently.

[0004] However, this type of technology has significant limitations. It relies on static data, making it difficult to integrate dynamic socioeconomic information such as population and transportation, and its understanding of demand-driven factors is insufficient. Furthermore, this current-state-based analytical approach lacks the ability to predict the dynamic evolution of future cities, resulting in insufficient foresight in planning schemes. More importantly, traditional methods tend to seek a single optimal solution, ignoring future uncertainties such as market and policy factors, making the planning schemes poorly adaptable in long-term operation and posing high investment risks.

[0005] Therefore, this invention proposes a smart site selection base station management system and method to address the shortcomings of existing technologies. Summary of the Invention

[0006] The purpose of this invention is to provide a smart site selection base station management system and method, which solves the problems of traditional site selection methods relying on static data and lacking dynamic prediction capabilities, resulting in insufficient foresight and robustness of planning schemes and high investment risks.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a smart site selection base station management system, comprising:

[0008] The urban multidimensional data acquisition and preprocessing module is used to collect and preprocess multidimensional urban data;

[0009] The knowledge base construction module builds a structured industry data knowledge base based on the multi-dimensional city data.

[0010] The urban evolution dynamics simulation module is used to simulate the evolution of urban dynamics and future charging demand based on data in the industry data knowledge base, through multi-agent modeling and spatiotemporal graph neural network-assisted calibration, and to generate corresponding simulation results.

[0011] The future scenario generation module is used to generate a scenario set containing multiple future urban evolution scenarios by setting different parameters and perturbations based on the simulation results.

[0012] The charging network optimization model construction module is used to analyze scenario characteristics based on the scenario set and define the optimization objective function and constraints for charging base station network planning in order to construct a charging base station network optimization model.

[0013] The robust optimization solution module is used to solve the charging base station network optimization model using a multi-objective robust optimization algorithm, and combined with a robustness metric, to generate a Pareto optimal robust charging base station network planning scheme set.

[0014] Preferably, the urban multidimensional data acquisition and preprocessing module includes:

[0015] The collection includes at least the following multi-dimensional urban data: geospatial data, population and socioeconomic data, traffic dynamics data, energy and infrastructure data, and urban planning and policy data.

[0016] The collected multi-dimensional urban data is subjected to data cleaning, data fusion, and data transformation processes.

[0017] Preferably, the knowledge base construction module includes:

[0018] Receive the multi-dimensional city data preprocessed by the city multi-dimensional data acquisition and preprocessing module;

[0019] The multi-dimensional urban data is then structured and spatiotemporally aligned.

[0020] The processed structured data is integrated and stored to form the structured industry data knowledge base.

[0021] Preferably, the urban evolution dynamics simulation module includes:

[0022] Based on data obtained from the industry data knowledge base, a virtual city environment containing geospatial information is constructed.

[0023] Define the types, attributes, behavioral rules, and decision-making logic of agents in the virtual city environment to perform multi-agent modeling;

[0024] The multi-agent model simulates the evolution of urban dynamics and future charging demand, and generates corresponding simulation results.

[0025] Preferably, when the urban evolution dynamics simulation module performs simulations using the multi-agent modeling, it employs a spatiotemporal graph neural network to assist in the calibration of the multi-agent modeling. The spatiotemporal graph neural network updates the node feature representations using the following formula. :

[0026] ;

[0027] In the formula, For the first Layer node feature representation matrix; The adjacency matrix for incorporating self-loops; for The angle matrix; For the first Layer-trainable weight matrix; This is the activation function.

[0028] Preferably, the future scenario generation module includes:

[0029] Receive the simulation results generated by the urban evolution dynamics simulation module;

[0030] Based on the simulation results, multiple sets of parameter combinations and perturbation factors were set;

[0031] Multiple rounds of urban evolution simulation are run to generate a scenario set containing various future urban evolution scenarios.

[0032] Preferably, when setting the multiple sets of parameter combinations and perturbation factors, the future scenario generation module, at least for some parameters, in the first... In the first scenario of future city evolution The setting value of each parameter It is one of the following:

[0033] or ;

[0034] In the formula, Representing the In the first scenario of future city evolution The setting values ​​of each parameter; Representing the The baseline values ​​for each parameter; Representative regarding the first The parameter in the first... Multiplicative perturbation factor in each scenario; Representative regarding the first The parameter in the first... Additive perturbation factor under each scenario;

[0035] The disturbance factor is sampled from a preset probability distribution or determined according to a predefined set of rules.

[0036] Preferably, the charging network optimization model construction module includes:

[0037] Based on the scenario set, the distribution of charging demand, traffic flow and spatiotemporal evolution trends under each scenario are analyzed to extract scenario features;

[0038] Define an optimization objective function that aims to minimize the total network cost and maximize service coverage;

[0039] Define constraints including capacity limits, total budget limits, and service level requirements for candidate sites;

[0040] Based on the aforementioned objective function and constraints, the charging base station network optimization model is constructed.

[0041] Preferably, the robust optimization solution module includes:

[0042] A multi-objective robust optimization algorithm, selected from the group consisting of an improved non-dominated sorting genetic algorithm and a multi-objective particle swarm optimization algorithm, is used to solve the charging base station network optimization model.

[0043] During the solution process, robustness metrics are combined to evaluate the performance stability of each planning scheme under different scenarios. The robustness metrics include regret value minimization metrics or worst-case performance optimization metrics.

[0044] The solution yields a set of solutions that balance the objective function and the robustness measure;

[0045] Identify and output the Pareto optimal robust charging base station network planning scheme set from the solutions.

[0046] This invention also provides a smart location-selection base station management method, comprising the following steps:

[0047] Collect and preprocess multi-dimensional urban data;

[0048] A structured industry data knowledge base is constructed based on the aforementioned multi-dimensional city data;

[0049] Based on the data in the industry data knowledge base, through multi-agent modeling and spatiotemporal graph neural network-assisted calibration, the evolution of urban dynamics and future charging demand is simulated, and corresponding simulation results are generated.

[0050] Based on the simulation results, a scenario set containing various future urban evolution scenarios is generated by setting different parameters and perturbations;

[0051] Based on the scenario set, the scenario characteristics are analyzed, and the optimization objective function and constraints for charging base station network planning are defined to construct a charging base station network optimization model.

[0052] Based on the aforementioned charging base station network optimization model, a multi-objective robust optimization algorithm is used to solve the problem, and a robustness metric is combined to generate a Pareto optimal robust charging base station network planning scheme set.

[0053] In summary, the present invention has at least one of the following beneficial technical effects:

[0054] 1. This invention overcomes the information limitations of traditional site selection methods that rely solely on static geographic information and expert experience by constructing a structured industry knowledge base encompassing multi-dimensional urban data such as population, transportation, and commerce. This solution integrates discrete and heterogeneous data into a machine-readable knowledge network, providing a comprehensive and timely digital foundation for subsequent dynamic simulations and intelligent decision-making. It significantly enhances the breadth and depth of data in intelligent site selection analysis, making the assessment of a location's commercial potential more accurate and reliable.

[0055] 2. This invention addresses the challenge of traditional site selection tools failing to effectively capture dynamic urban changes and predict future demand by introducing an urban evolution dynamics simulation module and combining multi-agent modeling with spatiotemporal graph neural network calibration technology. Instead of relying on static data analysis, this approach simulates the micro-behaviors and macro-emergence of various urban actors to extrapolate the dynamic evolution of future traffic flow and charging demand. This enables a forward-looking assessment of site selection decisions, significantly improving the timeliness and accuracy of site selection solutions.

[0056] 3. This invention addresses the lack of stability in traditional single-planning schemes when facing future uncertainties by constructing and solving a multi-objective robust optimization model, effectively reducing investment risk and decision-making costs. This scheme does not generate a single "optimal solution," but rather, by systematically considering multiple future evolution scenarios, generates a set of Pareto-optimal robust planning schemes that strike a balance between cost, coverage, and robustness. This allows decision-makers to select the most suitable scheme based on different strategic preferences, significantly improving the scientific rigor and decision-making flexibility of charging network planning. Attached Figure Description

[0057] Figure 1 This is a system architecture diagram of the present invention;

[0058] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0059] The following is in conjunction with the appendix Figure 1 -Appendix Figure 2The present invention will be further described in detail below.

[0060] This invention provides a smart site selection base station management system, comprising:

[0061] The urban multidimensional data acquisition and preprocessing module is used to collect and preprocess multidimensional urban data;

[0062] In this embodiment, the operational quality of the urban multidimensional data acquisition and preprocessing module directly determines the realism of the subsequent urban evolution simulation, the diversity of scenario generation, and the reliability and robustness of the final site selection scheme. Its internal workflow can be mainly divided into two closely linked stages: data acquisition and data preprocessing.

[0063] During the data acquisition phase, this module is configured to systematically acquire multi-dimensional urban data required for constructing a digital twin of the city from multiple heterogeneous information sources. Here, "multi-dimensional" refers not only to the diversity of data types but also to the temporal and spatial attributes. Preferably, the acquired multi-dimensional urban data types include at least:

[0064] Geospatial data: This forms the framework for constructing the virtual city environment. It includes not only the macroscopic urban road network topology and administrative boundaries, but also the microscopic land use types (such as commercial, residential, and industrial land) and the precise geographical locations of points of interest (POIs). The purpose of collecting this data is to provide a spatially realistic constraint for the subsequent activities of intelligent agents and to provide a foundation for calculating key indicators such as accessibility and service radius.

[0065] Demographic and socioeconomic data: This is the core driver of simulated charging demand. The module collects population density, age composition, income level, and vehicle ownership data for each Traffic Analysis Zone (TAZ) or a finer grid, especially the current penetration rate and growth forecast of electric vehicles (EVs). This data is crucial for initializing agent attributes, as residents from different socioeconomic backgrounds exhibit significant differences in travel patterns, charging habits, and willingness to pay.

[0066] Traffic dynamic data: This is crucial for reproducing the pulse of urban operations. The module obtains historical and real-time traffic flow, speed changes, and congestion indices for major road sections by accessing the city's traffic information center or parsing floating car data (FCD). Its purpose is not only to calibrate traffic flow in the simulation module but also to identify potential congestion bottlenecks, avoiding the planning of time-sensitive facilities such as fast-charging stations in these areas, thereby improving the user experience.

[0067] Energy and Facility Data: This serves as a physical constraint to ensure the feasibility of the planning scheme. The module collects the geographical location, rated capacity, and real-time load of each substation in the existing power grid. This is necessary because charging stations, especially DC fast charging stations, represent a significant load on the power grid, and their locations must have sufficient grid margin to avoid impacting the grid. Simultaneously, collecting data on the location and utilization rate of existing charging stations helps in competitive landscape analysis and identifying service blind spots.

[0068] Urban planning and policy data: This is key information that gives the model its forward-looking capabilities. The module collects and structures texts or documents such as the city's future master plan, regional development strategy, and new energy subsidy policies. Unlike other types of data that rely on historical data, this type of data reveals future development trends and is a necessary input for generating forward-thinking and robust planning solutions. It also guides the parameter settings of subsequent future scenario generation modules.

[0069] During the data preprocessing stage, this module performs a series of in-depth processing operations on the raw, heterogeneous, multi-dimensional urban data collected above, aiming to "purify" it into well-organized data assets that can be directly used by the model.

[0070] First, data cleaning is performed. For outliers, missing values, and noisy data in multi-dimensional urban data caused by sensor malfunctions, network transmission errors, or human input mistakes, professional statistical methods are used for processing. For example, time-series-based interpolation algorithms (such as linear interpolation and spline interpolation) can be used to fill in short-term gaps in traffic flow data; or anomaly detection algorithms such as isolated forests can be used to identify and remove obviously unreasonable data points.

[0071] Secondly, data fusion is performed. Since multi-dimensional urban data comes from a wide range of sources, its spatial references, temporal granularities, and data formats vary. The core task of this step is to achieve spatiotemporal alignment and format unification. Spatially, using Geographic Information System (GIS) spatial join technology, data from different layers (such as point-based POIs, linear road networks, and areal population areas) are mapped onto a unified geographic grid or traffic analysis area. Temporally, through resampling or aggregation, data of different frequencies (such as second-level vehicle trajectories and daily population statistics) are unified to a suitable temporal granularity for analysis, preferably hourly.

[0072] Finally, data transformation operations are performed. To meet the input data requirements of subsequent machine learning models (especially spatiotemporal graph neural networks), multi-dimensional urban data needs to be normalized. For example, for numerical features such as population size and traffic flow, min-max scaling or Z-score normalization methods are used to scale them to a fixed interval (such as [0,1] or a mean of 0 and a variance of 1) to eliminate dimensional differences and avoid gradient vanishing or exploding during model training. For categorical features such as land use type, one-hot encoding or similar methods are used to convert them into numerical vectors.

[0073] Through the above data collection and preprocessing process, this module ultimately outputs a clean, complete, multi-dimensional, and uniformly formatted structured urban data set.

[0074] The knowledge base construction module builds a structured industry data knowledge base based on the multi-dimensional city data.

[0075] In this embodiment, the main task of the knowledge base construction module is to receive the multi-dimensional city data processed by the preceding module and organize it into a structured industry data knowledge base containing internal connections.

[0076] This module is initially configured to receive multi-dimensional urban data preprocessed by the urban multi-dimensional data acquisition and preprocessing module. This means that the input data for this module already possesses a high degree of consistency and regularity, allowing the module to focus on building logical relationships between data and organizing knowledge.

[0077] After receiving the data, the module's key function—structuring and aligning multi-dimensional city data in time and space—begins to execute.

[0078] In terms of structured processing, considering that urban systems contain diverse elements and complex interactions between them, simple table structures have limitations in expressing this inherent correlation. Therefore, in a preferred embodiment, a heterogeneous information network (i.e., a graph structure) is used to organize urban data. Specifically, the module maps different types of urban entities (e.g., a point of interest, a land parcel, a road network intersection) as "nodes" in the network, and abstracts the relationships between entities (e.g., physical connections of road networks, spatial adjacency relationships between land parcels, commuting relationships from residential areas to work areas) as "edges" in the network. In this way, the originally relatively discrete multi-dimensional urban data is integrated into a knowledge graph that reflects the inherent logic.

[0079] Regarding spatiotemporal alignment, this step further processes the initial alignment from the previous module. Its purpose is to ensure that all nodes and edges in the graph structure have clearly defined and consistent spatiotemporal labels attached to their attributes. For example, the attributes of a road network "edge," in addition to static information such as length and grade, can also include a time series in hourly units to record its traffic flow at different times. Similarly, the attributes of a population area "node" can include population projection data for different years. This aims to ensure that the knowledge base can provide a logically consistent and state-consistent city snapshot at any time slice, providing a data consistency foundation for subsequent dynamic simulations.

[0080] Finally, the module integrates and stores the processed structured data to form a structured industry data knowledge base. This step instantiates and persistently stores the previously defined graph structure and its attached spatiotemporal attribute data. For storage technology selection, a graph database (such as Neo4j) is preferably used to store the city's topological network and semantic relationships to support efficient graph-related queries. Simultaneously, for geographic geometric information, a spatial database with spatiotemporal indexing capabilities (such as PostGIS) can be used to handle operations such as spatial extent queries.

[0081] Ultimately, the module outputs a structured, queryable, and analyzable industry data knowledge base.

[0082] The urban evolution dynamics simulation module is used to simulate the evolution of urban dynamics and future charging demand based on data in the industry data knowledge base, through multi-agent modeling and spatiotemporal graph neural network-assisted calibration, and to generate corresponding simulation results.

[0083] In this embodiment, the main task of the urban evolution dynamics simulation module is to use the existing industry data knowledge base to deduce the long-term dynamic evolution of the city in the future through a hybrid method that integrates micro-behavioral simulation and macro-data calibration, and generate corresponding spatiotemporal distribution results of future charging demand.

[0084] The operation of this module begins by constructing a virtual city environment containing geospatial information based on data obtained from an industry data knowledge base. This virtual city environment is not generated out of thin air, but rather a three-dimensional mapping of the structured data in the knowledge base. Specifically, the module reads geospatial information from the knowledge base, such as road network topology, land parcel functional divisions, and point of interest (POI) distribution, and constructs a digital city "sandbox" in computer memory. This sandbox forms the underlying spatial foundation for all simulation activities.

[0085] Subsequently, within this virtual city environment, the module defines the types, attributes, behavioral rules, and decision-making logic of agents to perform multi-agent modeling. Here, an agent is a computational abstraction of various decision-making entities in the city.

[0086] In a specific implementation, the types of intelligent agents can include residents, commuter vehicles, business entities, etc.

[0087] Its attributes are initialized from the corresponding demographic and socioeconomic data in the knowledge base. For example, attributes such as age, income, place of residence, place of work, and whether or not a resident agent owns an electric vehicle are assigned to the resident agent.

[0088] Its behavioral rules define the daily activity patterns of intelligent agents, such as travel chains for commuting, shopping, and leisure based on historical travel survey (OD) data.

[0089] Its decision-making logic is even more crucial, as it defines how an intelligent agent behaves when faced with choices. For example, when the battery level of an electric vehicle owner is below a certain threshold, the intelligent agent will decide which charging station to go to based on factors such as the distance to the destination, the expected charging cost, and the queuing time at the charging station.

[0090] After completing the above definition, the module simulates the evolution of urban dynamics and future charging demand through multi-agent modeling. Tens of thousands of agents interact in parallel in the virtual environment. Their micro-behaviors (such as individual travel and charging choices) come together to reveal dynamic phenomena of the entire city at the macro level, such as the spatiotemporal distribution of traffic flow, the periodic prosperity of commercial areas, and most importantly, the future charging demand of different areas at different times.

[0091] However, simple multi-agent models, due to their numerous parameters, may produce simulation results that deviate from the real world. To improve the fidelity of the simulation, a key technical feature of this module is the use of a spatiotemporal graphical neural network (ST-GNN) to assist in the calibration of the multi-agent model. This is necessary because ST-GNN can learn the complex and nonlinear spatiotemporal dependencies of the urban system from a large amount of historical spatiotemporal data (such as traffic flow and charging records) stored in a knowledge base. These learned macroscopic patterns are used to constrain and calibrate the microscopic behavioral parameters in the multi-agent model, ensuring that the simulated macroscopic phenomena are consistent with the statistical characteristics of real data.

[0092] In a preferred embodiment, the spatiotemporal graph neural network updates the node feature representations using the following formula. :

[0093] ;

[0094] This formula describes a core operation in graph convolutional networks, which aggregates feature information from neighboring nodes in the graph.

[0095] In the formula, For the first The node feature representation matrix of the layer, where each row can represent a feature vector of a region at a specific time (such as population, traffic flow, etc.). The adjacency matrix with self-loops is used to represent the spatial connectivity between regions; for The angle matrix is ​​used for normalization to maintain data scale stability; For the first The layer is a trainable weight matrix that encodes the patterns of how spatiotemporal features are transformed and transmitted. The activation function (such as ReLU) enables the model to learn and express complex nonlinear relationships.

[0096] Through this mechanism, the simulation results generated by this module not only stem from the deduction of microscopic behavior, but are also calibrated by the patterns of macroscopic data.

[0097] The future scenario generation module is used to generate a scenario set containing multiple future urban evolution scenarios by setting different parameters and perturbations based on the simulation results.

[0098] In this embodiment, the future scenario generation module is a core component for addressing profound uncertainties about the future. Its main purpose is not to make a single, precise prediction, but to explore multiple possible future development paths through a systematic approach, thereby constructing a set of scenarios for robust evaluation and optimization.

[0099] The module's operation begins with receiving simulation results generated by the Urban Evolution Dynamics Simulation Module. These simulation results typically represent a "baseline" or "business as usual" future development path—the most probable outcome based on current understanding. However, relying solely on this single scenario for long-term planning carries significant risk.

[0100] Therefore, the core function of the module is to explore a broad range of possibilities beyond the baseline scenario by setting multiple combinations of parameters and perturbation factors based on simulation results. This process first requires identifying key uncertainty parameters that have a significant impact on the long-term evolution of the city and charging demand. Preferably, these parameters may include macroeconomic growth rate, the rate of progress and cost reduction curve of electric vehicle technology, long-term fluctuations in fuel prices, the continuity and intensity of relevant subsidy policies, and urban spatial expansion patterns.

[0101] This module employs a structured perturbation method when setting parameters. In a preferred embodiment, at least for some parameters, in the... In the first scenario of future city evolution The setting value of each parameter It is one of the following:

[0102] or ;

[0103] The formula here provides a standardized approach to generating new parameter values ​​from a baseline.

[0104] In the formula, Representing the In the first scenario of future city evolution The setting values ​​of each parameter; Representing the The baseline values ​​for each parameter; Representative regarding the first The parameter in the first... Multiplicative perturbation factor in each scenario; Representative regarding the first The parameter in the first... Additive perturbation factors under each scenario.

[0105] The first form (multiplicative perturbation) uses the perturbation factor. This is applicable to parameters describing relative rates of change, such as "economic growth rate is 20% higher than the benchmark".

[0106] The second form (additive perturbation) uses a perturbation factor.

[0107] Implementation, suitable for describing changes in absolute values.

[0108] Disturbance factor and The source of this information is key to enabling scenario diversity in this module. In a specific implementation, the perturbation factor is sampled from a preset probability distribution or determined according to a predefined set of rules.

[0109] When sampling from probability distributions, for example, one can assume that the disturbance factor of the economic growth rate follows a normal distribution with a mean of 0. This means that in most scenarios, economic growth will fluctuate slightly around the baseline value, but there is also a small probability of significant deviations. This method can systematically and on a large scale generate statistically significant scenarios.

[0110] When determined based on a set of rules, scenarios with a clear narrative logic can be constructed using expert knowledge or policy analysis. For example, a "radical transition" scenario can be defined, which simultaneously sets a combination of parameters such as large subsidies for electric vehicles, high fuel prices, and rapid technological progress. This approach ensures a thorough assessment of certain key, structural future possibilities.

[0111] After setting a complete set of perturbation combinations for all key parameters, the module executes multiple rounds of urban evolution simulation. Each parameter combination is substituted into the preceding "Urban Evolution Dynamics Simulation Module" and a complete simulation from the present to the future is run. The output of each round constitutes an independent future urban evolution scenario, containing details of the spatiotemporal distribution of future charging demand under that parameter setting.

[0112] By repeating this process, the module eventually generates a scenario set containing various future urban evolution scenarios.

[0113] The charging network optimization model construction module is used to analyze scenario characteristics based on the scenario set and define the optimization objective function and constraints for charging base station network planning in order to construct a charging base station network optimization model.

[0114] In this embodiment, the core function of the charging network optimization model construction module is to transform the complex, multi-dimensional charging base station network planning problem into a structured, solvable mathematical optimization model. This module serves as a bridge connecting "scenario generation" and "optimization solution," and is responsible for formally expressing the various future possibilities generated by the preceding modules as the objectives and constraints of the optimization problem.

[0115] This module begins by analyzing the distribution of charging demand, traffic flow, and spatiotemporal evolution trends under a scenario set to extract scenario features. It receives a scenario set output by the "Future Scenario Generation Module" and analyzes each scenario. This analysis aims to quantify the performance potential of each candidate site under different future scenarios. For example, the module calculates the potential charging demand within the service area of ​​each candidate site under each scenario, as well as key indicators such as peak-hour traffic flow, providing specific, quantitative data input for subsequently defining the objective function and constraints.

[0116] Next, the module enters the core mathematical modeling phase, which begins by defining an optimization objective function that aims to minimize at least one of the total network cost and maximize service coverage. In a preferred embodiment, this is a multi-objective optimization problem to balance the economy and service effectiveness of the planning scheme.

[0117] Objective 1: Minimize the total network cost. This objective focuses on the economic feasibility of the proposed solution, aiming to minimize the sum of the costs associated with all selected candidate sites. Its mathematical form can be expressed as:

[0118] ;

[0119] In the formula, Represents the set of all candidate sites; This is a binary decision variable; a value of 1 indicates that the candidate position is... Build charging stations; if the number is 0, do not build them. Represents the position The total cost associated with website construction may include the discounted costs of construction, land acquisition, and operation and maintenance. Let be the objective function value of the total network cost to be optimized.

[0120] Objective Two: Maximize service coverage. This objective focuses on the service quality and robustness of the solution, aiming to improve performance in the worst-case scenario across all possible scenarios. Its mathematical form can be expressed as:

[0121] ;

[0122] In the formula, A set representing all future scenarios; Represents the set of all charging demand points; Representative demand points In the context The charging demand under; As an auxiliary binary variable, its value of 1 represents the demand point. In the context The lower part can be covered by the network; Let be the objective function value for robust service coverage to be optimized. This formula reflects a consideration of the performance stability of the planning scheme by maximizing the minimum service coverage across all scenarios.

[0123] Subsequently, the module definition includes constraints such as capacity limits, total budget limits, and service level requirements for candidate sites. These constraints ensure the feasibility of the final solution in the real world.

[0124] Total budget limit: This constraint ensures that the total cost of the plan will not exceed the preset total investment budget. . ;

[0125] Candidate site capacity constraints: This constraint ensures that the total demand served by each planned charging station will not exceed its design capacity in any future scenario. .

[0126] ;

[0127] In the formula, For the context Sub-assigned to sites The total charging demand. This constraint must hold for every candidate site and every future scenario.

[0128] Service Level Requirements: This is the bottom-line constraint for service quality, requiring the network to meet a minimum service coverage percentage under any future scenario. .

[0129] ;

[0130] This constraint also needs to be satisfied for every future scenario.

[0131] Finally, this module comprehensively optimizes the objective function and constraints to construct a charging base station network optimization model. This model mathematically describes the complex decision-making problem of "finding a series of charging network layout schemes that achieve different trade-offs between cost and service coverage, while satisfying realistic constraints such as budget, capacity, and service level, and ensuring that these trade-offs remain as robust as possible under various future scenarios."

[0132] The robust optimization solution module is used to solve the charging base station network optimization model using a multi-objective robust optimization algorithm, and combined with a robustness metric, to generate a Pareto optimal robust charging base station network planning scheme set.

[0133] In this embodiment, the core task of the robust optimization solution module is to efficiently solve the charging base station network optimization model constructed by the preceding module, which contains multiple objectives and future uncertainties, and ultimately generate a series of planning schemes that achieve a trade-off between economy, serviceability and robustness.

[0134] This module first receives a mathematical model defined by the "Charging Network Optimization Model Construction Module," which includes a complete objective function and constraints. Then, it employs a multi-objective robust optimization algorithm selected from a group consisting of an improved non-dominated sorting genetic algorithm and a multi-objective particle swarm optimization algorithm to solve the charging base station network optimization model. This type of heuristic algorithm is preferred because the charging network location problem is typically large-scale, multi-constraint, and multi-objective, and the objective function may be non-convex, making it difficult for traditional exact algorithms to solve in a reasonable time. Improved non-dominated sorting genetic algorithms (such as NSGA-II), by simulating selection, crossover, and mutation in biological evolution, can perform a global search in a complex solution space, making them particularly suitable for finding the Pareto optimal solution set for multi-objective problems.

[0135] A key technical feature of this module during the solution process is the integration of robustness metrics to evaluate the performance stability of each planning scheme under different scenarios. This aims to ensure that the generated scheme not only performs well under a desired scenario but also can withstand the negative impacts of future uncertainties. Preferably, the robustness metrics include a regret minimization metric or a worst-case performance optimization metric.

[0136] Worst-case performance optimization metric: This metric was formally defined as the robust service coverage objective function in the previous "Charging Network Optimization Model Construction Module".

[0137] This solution module directly minimizes the total network cost when executing the algorithm.

[0138] And maximizing

[0139] As two core optimization objectives, the algorithm's search process converges towards lower cost and higher baseline performance.

[0140] Regret minimization metric: This is an alternative or robustness evaluation method that can be used in conjunction with the former. It measures the performance gap between a solution's performance in a given scenario and the "optimal solution designed specifically for that scenario if it could be predicted." This gap is the "regret value." The goal of this metric is to find a solution that minimizes the "maximum regret value" across all possible scenarios. Its mathematical form can be defined as:

[0141] ;

[0142] In the formula, This represents the planning options currently being evaluated. Represents a specific future scenario; It is a plan In the context Performance values ​​(such as service coverage) under these conditions; Represents the known situation Given that this scenario will occur, the theoretically optimal solution that achieves best performance under this condition is... This is the maximum regret value of the proposed solution. When using this metric, it can be incorporated into the algorithm as a third optimization objective.

[0143] In each iteration of the algorithm, for each candidate solution (i.e., a specific charging network planning scheme) in the population, the module calculates its cost target value. And one or more of the aforementioned robustness measures, and use these as the basis for non-dominated sorting and selection operations.

[0144] Through multi-generational evolutionary computation, the module ultimately obtains a set of solutions that balance the objective function and robustness metrics. These solutions are non-dominant to each other in the solution space.

[0145] Finally, the module identifies and outputs a set of Pareto-optimal robust charging station network planning schemes from the solutions. This set of schemes constitutes the final decision support outcome. Each scheme in the set is a Pareto-optimal solution, meaning that it is difficult to improve another objective without sacrificing at least one objective (such as increasing costs or reducing robustness). This set of schemes provides decision-makers with a series of high-quality alternatives, enabling them to select the most suitable charging network planning scheme based on their risk preferences and strategic priorities.

[0146] This invention also provides a smart location-selection base station management method, comprising the following steps:

[0147] S1. Collect and preprocess multi-dimensional city data;

[0148] S2. Construct a structured industry data knowledge base based on the multi-dimensional city data;

[0149] S3. Based on the data in the industry data knowledge base, through multi-agent modeling and spatiotemporal graph neural network-assisted calibration, simulate the evolution of urban dynamics and future charging demand, and generate corresponding simulation results.

[0150] S4. Based on the simulation results, by setting different parameters and perturbations, a scenario set containing multiple future urban evolution scenarios is generated;

[0151] S5. Based on the scenario set, analyze the scenario characteristics and define the optimization objective function and constraints for charging base station network planning to construct a charging base station network optimization model.

[0152] S6. Based on the aforementioned charging base station network optimization model, a multi-objective robust optimization algorithm is used to solve the problem, and a robustness metric is combined to generate a Pareto optimal robust charging base station network planning scheme set.

[0153] The method in this embodiment can be used to execute the above system embodiment, and its principle and technical effect are similar, so it will not be described again here.

[0154] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart site selection base station management system, characterized in that, include: The urban multidimensional data acquisition and preprocessing module is used to collect and preprocess multidimensional urban data; The knowledge base construction module builds a structured industry data knowledge base based on the multi-dimensional city data. The urban evolution dynamics simulation module is used to simulate the evolution of urban dynamics and future charging demand based on data in the industry data knowledge base, through multi-agent modeling and spatiotemporal graph neural network-assisted calibration, and to generate corresponding simulation results. The future scenario generation module is used to generate a scenario set containing multiple future urban evolution scenarios by setting different parameters and perturbations based on the simulation results. The charging network optimization model construction module is used to analyze scenario characteristics based on the scenario set and define the optimization objective function and constraints for charging base station network planning in order to construct a charging base station network optimization model. The robust optimization solution module is used to solve the charging base station network optimization model using a multi-objective robust optimization algorithm, and combined with a robustness metric, to generate a Pareto optimal robust charging base station network planning scheme set.

2. The intelligent site selection base station management system according to claim 1, characterized in that, The urban multidimensional data acquisition and preprocessing module includes: The collection includes at least the following multi-dimensional urban data: geospatial data, population and socioeconomic data, traffic dynamics data, energy and infrastructure data, and urban planning and policy data. The collected multi-dimensional urban data is subjected to data cleaning, data fusion, and data transformation processes.

3. The intelligent site selection base station management system according to claim 1, characterized in that, The knowledge base construction module includes: Receive the multi-dimensional city data preprocessed by the city multi-dimensional data acquisition and preprocessing module; The multi-dimensional urban data is then structured and spatiotemporally aligned. The processed structured data is integrated and stored to form the structured industry data knowledge base.

4. The intelligent site selection base station management system according to claim 1, characterized in that, The urban evolution dynamics simulation module includes: Based on data obtained from the industry data knowledge base, a virtual city environment containing geospatial information is constructed. Define the types, attributes, behavioral rules, and decision-making logic of agents in the virtual city environment to perform multi-agent modeling; The multi-agent model simulates the evolution of urban dynamics and future charging demand, and generates corresponding simulation results.

5. The intelligent site selection base station management system according to claim 4, characterized in that, When simulating urban evolution dynamics through multi-agent modeling, the urban evolution dynamics simulation module employs a spatiotemporal graph neural network to assist in the calibration of the multi-agent modeling. The spatiotemporal graph neural network updates the node feature representations using the following formula. : ； In the formula, For the first Layer node feature representation matrix; The adjacency matrix for incorporating self-loops; for The angle matrix; For the first Layer-trainable weight matrix; This is the activation function.

6. The intelligent site selection base station management system according to claim 1, characterized in that, The future scenario generation module includes: Receive the simulation results generated by the urban evolution dynamics simulation module; Based on the simulation results, multiple sets of parameter combinations and perturbation factors were set; Multiple rounds of urban evolution simulation are run to generate a scenario set containing various future urban evolution scenarios.

7. The intelligent site selection base station management system according to claim 6, characterized in that, When setting the multiple sets of parameter combinations and perturbation factors, the future scenario generation module, at least for some parameters, in the first... In the first scenario of future city evolution The setting value of each parameter It is one of the following: or ; In the formula, Representing the In the first scenario of future city evolution The setting values ​​of each parameter; Representing the The baseline values ​​for each parameter; Representative regarding the first The parameter in the first... Multiplicative perturbation factor in each scenario; Representative regarding the first The parameter in the first... Additive perturbation factor under each scenario; The disturbance factor is sampled from a preset probability distribution or determined according to a predefined set of rules.

8. The intelligent site selection base station management system according to claim 1, characterized in that, The charging network optimization model construction module includes: Based on the scenario set, the distribution of charging demand, traffic flow and spatiotemporal evolution trends under each scenario are analyzed to extract scenario features; Define an optimization objective function that aims to minimize the total network cost and maximize service coverage; Define constraints including capacity limits, total budget limits, and service level requirements for candidate sites; Based on the aforementioned objective function and constraints, the charging base station network optimization model is constructed.

9. The intelligent site selection base station management system according to claim 1, characterized in that, The robust optimization solution module includes: A multi-objective robust optimization algorithm, selected from the group consisting of an improved non-dominated sorting genetic algorithm and a multi-objective particle swarm optimization algorithm, is used to solve the charging base station network optimization model. During the solution process, robustness metrics are combined to evaluate the performance stability of each planning scheme under different scenarios. The robustness metrics include regret value minimization metrics or worst-case performance optimization metrics. The solution yields a set of solutions that balance the objective function and the robustness measure; Identify and output the Pareto optimal robust charging base station network planning scheme set from the solutions.

10. A smart location-selection base station management method, applied to the smart location-selection base station management system as described in any one of claims 1-9, characterized in that, Includes the following steps: Collect and preprocess multi-dimensional urban data; A structured industry data knowledge base is constructed based on the aforementioned multi-dimensional city data; Based on the data in the industry data knowledge base, through multi-agent modeling and spatiotemporal graph neural network-assisted calibration, the evolution of urban dynamics and future charging demand is simulated, and corresponding simulation results are generated. Based on the simulation results, a scenario set containing various future urban evolution scenarios is generated by setting different parameters and perturbations; Based on the scenario set, the scenario characteristics are analyzed, and the optimization objective function and constraints for charging base station network planning are defined to construct a charging base station network optimization model. Based on the aforementioned charging base station network optimization model, a multi-objective robust optimization algorithm is used to solve the problem, and a robustness metric is combined to generate a Pareto optimal robust charging base station network planning scheme set.

Images