A main micro-configuration synergistic electric vehicle load potential evaluation method, system and medium

Abstract

This invention relates to the field of power grid coordinated control technology, and particularly to a method, system, and medium for assessing the load potential of electric vehicles through a coordinated approach between the main grid, distribution network, and microgrid. The method includes: acquiring power grid topology information, charging behavior data, user travel data, and power grid operation data; forming grid cells from the distribution network; constructing a coordinated architecture of the main grid layer, distribution network layer, and microgrid layer based on the grid cells; generating user behavior profiles based on the charging behavior data and user travel data; predicting the charging load of each grid cell based on the user behavior profiles and power grid operation data; aggregating the adjustable power of the electric vehicle group, constructing a generalized energy storage equivalent model, and determining the regulation potential boundary; generating a guidance strategy and outputting charging load adjustment instructions, and continuously correcting the electric vehicle charging load and regulation potential boundary. This invention effectively solves the problems of inaccurate electric vehicle load prediction and distorted regulation potential assessment, which make it difficult to support refined power grid control.

Description

Technical Field

[0001] This invention relates to the field of power grid coordinated control technology, and in particular to a method, system and medium for assessing the load potential of electric vehicles through main distribution and micro-coordination. Background Technology

[0002] With the large-scale integration of new energy vehicles into the power distribution network, electric vehicle charging load has gradually become an important factor affecting the safe operation and flexible regulation of the power grid. Among them, electric vehicle load forecasting refers to the prediction of future charging demand by combining vehicle charging behavior, travel patterns and power grid operation status. Adjustable potential assessment refers to the assessment of the actual adjustment capacity of the electric vehicle group for peak shaving, demand response and new energy consumption under the premise of meeting user travel needs and battery operation constraints. Existing technologies typically employ load forecasting methods based on statistical laws or single models, and make overall estimates based on substations, feeders, or individual sites. Alternatively, they may conduct static assessments of the adjustable resources of electric vehicles based on idealized assumptions. While these methods can reflect the changing trends and adjustment capabilities of charging loads to some extent, they generally lack a unified modeling of the collaborative relationships between the main grid, distribution network, and microgrids. They fail to fully consider differences in the spatial granularity of the distribution network, regional functional attributes, and behavioral differences such as user travel habits, charging preferences, and response intentions. Furthermore, they lack a dynamic correction mechanism that combines real-time access status and historical response results. Consequently, the accuracy of electric vehicle charging load forecasting is insufficient, and the assessment of adjustment potential is prone to distortion. These methods fail to accurately reflect the adjustability of electric vehicle groups under the constraints of different regions and levels of the power grid, thus hindering effective support for refined power grid control and vehicle-grid interaction.

[0003] The information disclosed in this background section is intended only to enhance the understanding of the general background of this disclosure and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0004] This invention provides a method, system, and medium for assessing the load potential of a primary-secondary-micro cooperative electric vehicle, which can effectively solve the problems in the background art.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for assessing the load potential of a primary-secondary-micro cooperative electric vehicle, the method comprising: The system acquires power grid topology information, charging behavior data, user travel data, power grid operation data, and regional functional attribute data for the target area. Based on the power grid topology information and the regional functional attribute data, the distribution network is divided into grids to form dynamically adjustable grid units. Based on the grid units and the power grid operation data, a multi-level power grid collaborative architecture is constructed, which coordinates the main grid layer, distribution network layer, and microgrid layer, and a mapping relationship is established between the grid units and the upper-level distribution facilities. Based on the charging behavior data and the user travel data, the travel characteristics, charging preferences and response characteristics of electric vehicle users within the grid unit are extracted to generate a user behavior profile corresponding to the grid unit; Based on the user behavior profile, the multi-level power grid collaborative architecture, and the power grid operation data, predict the electric vehicle charging load of each grid unit during the target time period; Based on the electric vehicle charging load, the user behavior profile, and the multi-level power grid collaborative architecture, the adjustable power of the electric vehicle group within the grid unit is aggregated to construct a generalized energy storage equivalent model and determine the adjustment potential boundary of each grid unit. Based on the regulation potential boundary and the regulation requirements corresponding to the multi-level power grid collaborative architecture, a differentiated guidance strategy is generated, and a charging load regulation command is output to the corresponding grid unit. The electric vehicle charging load and the regulation potential boundary are then rolled over according to the real-time vehicle access status and historical response results.

[0006] Furthermore, the distribution network is divided into grids, including: Acquire geographic information system data, power distribution network topology data, and point of interest (POI) data for the target area; The grid cell boundaries are determined based on the geographic information system data, the power distribution network topology data, the regional load density, the power supply radius, and the point of interest (POI) data. The boundaries of the grid cells are dynamically adjusted based on changes in load hotspots.

[0007] Furthermore, establishing the mapping relationship between the grid unit and the upper-level power distribution facility includes: A collaborative operation framework consisting of a main grid layer, a distribution grid layer, and a microgrid layer is constructed based on the aforementioned grid units. The main grid layer is designated for energy market trading and system-level frequency regulation and peak shaving, the distribution grid layer is designated for feeder load balancing, voltage quality control and peak shaving and valley filling, and the microgrid layer is designated for local renewable energy consumption, islanded operation and demand response. Establish hierarchical transmission paths between the grid cells, feeders, substations, and main grid nodes.

[0008] Further, generating a user behavior profile corresponding to the grid cell includes: Based on historical charging records and travel data analysis, the travel chain patterns of electric vehicle users within the grid unit are analyzed. The travel time distribution, charging preferences, state of charge (SOC) trigger threshold, and target charging SOC level of the electric vehicle users are obtained. The electric vehicle users are differentiated into groups using a clustering algorithm, and the user behavior profile corresponding to the grid unit is generated by combining the functional attributes of the grid unit.

[0009] Furthermore, predicting the electric vehicle charging load of each grid cell during the target time period includes: Establish a travel transfer probability model for the electric vehicle users arriving at the grid cell in the future; The potential charging demand of each grid unit is generated by integrating the user behavior profile and the grid regulation signal corresponding to the multi-level power grid collaborative architecture. The predicted electric vehicle charging load for each grid cell within the target time period is output using a long short-term memory network or a Transformer model.

[0010] Furthermore, a generalized energy storage equivalent model is constructed and the regulation potential boundary of each grid cell is determined, including: Based on the current state of charge, maximum charging power, and expected off-grid time of each electric vehicle in the grid cell, a set of available charging and discharging capabilities is constructed. The public charging stations within the grid unit are considered as aggregated generalized energy storage units. The Minkowski summation method is used to vector-superimpose the adjustable power ranges of N electric vehicles within the grid cell over the next T time periods to obtain the total adjustable power envelope.

[0011] Further, determining the adjustment potential boundary of each of the grid cells includes: Construct a regulation potential index system, which includes at least transferable power, adjustable power, response speed, and reliability coefficient. The mechanistic constraint boundary of the grid cell is determined based on battery charging and discharging characteristics, vehicle range constraints, and users' minimum travel requirements. The user response rate prediction model is trained based on historical response data, and the adjustment potential boundary of the grid cell is determined by combining the mechanistic constraint boundary.

[0012] Furthermore, the electric vehicle charging load and the adjustment potential boundary are rolled over based on real-time vehicle access status and historical response results, including: The regulation requirements of the main grid layer, the distribution grid layer, and the microgrid layer are matched according to the regulation potential boundary; Generate a differentiated guidance strategy that includes at least one of electricity price incentives, credit points, and priority service rights, and output a charging load adjustment command to the corresponding grid cell; Based on the previous forecast, the electric vehicle charging load and the adjustment potential boundary are updated according to the latest vehicle access status at a preset rolling period of 15 minutes.

[0013] A load potential assessment system for a primary-secondary-micro cooperative electric vehicle, the system comprising: The data acquisition module acquires the power grid topology information, charging behavior data, user travel data, power grid operation data, and regional functional attribute data of the target area. Based on the power grid topology information and regional functional attribute data, the distribution network is divided into grids to form dynamically adjustable grid units. The mapping construction module constructs a multi-level power grid collaborative architecture that integrates the main grid layer, distribution network layer, and microgrid layer based on grid cells and power grid operation data, and establishes the mapping relationship between grid cells and upper-level distribution facilities. The profile generation module extracts the travel characteristics, charging preferences, and response characteristics of electric vehicle users within the grid unit based on charging behavior data and user travel data, in order to generate user behavior profiles corresponding to the grid unit. The load forecasting module predicts the electric vehicle charging load of each grid unit within the target time period based on user behavior profiles, multi-level power grid collaborative architecture, and power grid operation data. The boundary adjustment module aggregates the adjustable power of electric vehicle groups within a grid unit based on electric vehicle charging load, user behavior profiles, and a multi-level power grid collaborative architecture, in order to construct a generalized energy storage equivalent model and determine the adjustment potential boundary of each grid unit. The instruction regulation module generates differentiated guidance strategies based on the regulation potential boundary and the regulation requirements corresponding to the multi-level power grid collaborative architecture, and outputs charging load regulation instructions to the corresponding grid units. It also performs rolling corrections on the electric vehicle charging load and regulation potential boundary based on the real-time vehicle access status and historical response results.

[0014] A computer-readable storage medium storing a computer program, the computer program including program instructions that, when executed by a processor, can implement the aforementioned method for assessing the load potential of a primary-secondary-micro cooperative electric vehicle.

[0015] The technical solution of this invention can achieve the following technical effects: By acquiring grid topology, multi-source operation data, and user behavior data, the distribution network is divided into grids and a main-distribution-micro collaborative architecture is constructed. Then, combined with user behavior profiles, grid-level charging load forecasting, generalized energy storage aggregation, and regulation potential assessment are carried out. Finally, differentiated guidance strategies are generated and closed-loop control is achieved through rolling corrections. This effectively solves the problems of inaccurate electric vehicle load forecasting and distorted regulation potential assessment that make it difficult to support refined grid control.

[0016] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating a method for assessing the load potential of a primary-support-micro cooperative electric vehicle. Figure 2 This is the overall architecture diagram corresponding to Example 4; Figure 3 A comparison chart of the main grid's current power level and the actual power boundary. Figure 4 The distribution network power versus state of charge (SOC) curves; Figure 5 This is a comparison diagram of the microgrid's current boundary with its actual boundary. Figure 6 Create a flowchart for the functional area grid. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0021] Example 1; like Figure 1 As shown, this application provides a method for assessing the load potential of a primary-secondary micro-cooperative electric vehicle, the method comprising: S10: Obtain the power grid topology information, charging behavior data, user travel data, power grid operation data and regional functional attribute data of the target area, and divide the distribution network into grids based on the power grid topology information and regional functional attribute data to form dynamically adjustable grid units. S20: Based on grid units and power grid operation data, construct a multi-level power grid collaborative architecture that coordinates the main grid layer, distribution network layer, and microgrid layer, and establish a mapping relationship between grid units and upper-level distribution facilities; S30: Extract the travel characteristics, charging preferences and response characteristics of electric vehicle users within the grid unit based on charging behavior data and user travel data, so as to generate user behavior profiles corresponding to the grid unit; S40: Based on user behavior profiles, multi-level power grid collaborative architecture, and power grid operation data, predict the electric vehicle charging load of each grid unit during the target time period; S50: Based on electric vehicle charging load, user behavior profiles, and the adjustable power of electric vehicle groups within a multi-level grid collaborative architecture, a generalized energy storage equivalent model is constructed to determine the adjustment potential boundary of each grid unit. S60: Generate differentiated guidance strategies based on the regulation potential boundary and the regulation requirements corresponding to the multi-level power grid collaborative architecture, and output charging load regulation instructions to the corresponding grid units. Also, perform rolling corrections on the electric vehicle charging load and regulation potential boundary based on the real-time vehicle access status and historical response results.

[0022] Specifically, firstly, the system acquires grid topology information, charging behavior data, user travel data, grid operation data, and regional functional attribute data for the target area, and cleans, aligns, and preprocesses the data. Based on the grid topology information and regional functional attribute data, the distribution network is divided into grids, forming dynamically adjustable grid units, and each charging event is matched to the corresponding grid unit. Secondly, based on the grid units and grid operation data, a multi-level grid collaborative architecture is constructed, integrating the main grid layer, distribution network layer, and microgrid layer, establishing a mapping relationship between grid units and upper-level distribution facilities to realize the transmission of grid control demands at different levels. Thirdly, based on charging behavior data and user travel data, the system extracts the travel characteristics, charging preferences, and response characteristics of electric vehicle users within the grid units, and generates corresponding user behavior profiles by combining the functional attributes of the grid units. Finally, based on the user behavior profiles, the multi-level grid collaborative architecture, and grid operation data, the system predicts the electric vehicle charging load of each grid unit during the target time period.

[0023] After obtaining the electric vehicle charging load of each grid unit, the adjustable power of the electric vehicle group within the grid unit is aggregated based on the electric vehicle charging load, user behavior profiles, and multi-level grid collaborative architecture to construct a generalized energy storage equivalent model and determine the regulation potential boundary of each grid unit. The determination of the regulation potential boundary comprehensively considers battery characteristics, user travel demand, and historical response data. Subsequently, differentiated guidance strategies are generated based on the regulation potential boundary and the corresponding regulation requirements of the multi-level grid collaborative architecture, and charging load regulation instructions are output to the corresponding grid units. Preferably, the electric vehicle charging load and regulation potential boundary are rolled over according to real-time vehicle access status and historical response results to form a closed-loop regulation process, improving the accuracy of electric vehicle charging load prediction and regulation potential assessment.

[0024] The technical solution of this invention acquires power grid topology, multi-source operation data, and user behavior data, divides the distribution network into grids, and constructs a main-distribution-micro collaborative architecture. Then, combined with user behavior profiles, it conducts grid unit-level charging load prediction, generalized energy storage aggregation, and regulation potential assessment. Finally, it generates differentiated guidance strategies and achieves closed-loop control through rolling corrections, effectively solving the problems of inaccurate electric vehicle load prediction and distorted regulation potential assessment that make it difficult to support refined power grid control.

[0025] Furthermore, the distribution network is divided into grids, including: Acquire geographic information system data, power distribution network topology data, and point of interest (POI) data for the target area; The grid cell boundaries are determined based on geographic information system data, distribution network topology data, regional load density, power supply radius, and point of interest (POI) data. The boundaries of the grid cells are dynamically adjusted based on changes in load hotspots.

[0026] As a preferred embodiment of the above embodiments, in one embodiment, the distribution network is divided into grids based on the power grid topology information and regional functional attribute data to form dynamically adjustable grid units. Specifically, this can be achieved in the following way: First, the geographic information system map of the target area, the distribution network topology data, and the point of interest (POI) information are obtained. The geographic information system map is used to represent the spatial location relationship of the target area, the distribution network topology data is used to represent the line connection relationship, power supply range, and equipment distribution, and the point of interest (POI) information is used to represent the functional attributes of different areas. After obtaining the above data, the distribution network is divided into several grids based on the load distribution characteristics, power supply radius, and functional zone attributes within the region. Each grid unit corresponds to a certain geographical area and has relatively independent power supply characteristics, serving as the spatial basis unit for subsequent electric vehicle charging demand prediction and regulation. In specific implementation, geographical topology, load density, power supply radius, and point of interest (POI) functional attributes can be comprehensively considered to divide the boundaries of different areas, so as to avoid the failure of travel trajectory aggregation due to excessively large spatial scale, or the increase of empty trajectory grids and model calculation complexity due to excessively small spatial scale. Furthermore, after the grid division is completed, each charging event can be matched to the corresponding grid cell based on the geographic location, and the functional attributes of each grid cell can be determined by combining the point of interest (POI) data to achieve spatial attribution labeling of charging events. Preferably, the boundary of the grid cell is not fixed, but can be dynamically adjusted according to the changes in load hotspots, thereby supporting differentiated modeling in different functional areas of the city and improving the spatial resolution and feasibility of subsequent load forecasting and adjustment potential assessment.

[0027] Furthermore, establishing a mapping relationship between grid cells and upper-level power distribution facilities includes: A collaborative operation framework consisting of a main grid layer, a distribution grid layer, and a microgrid layer is constructed based on grid cells; The main grid layer is designated for energy market trading and system-level frequency regulation and peak shaving; the distribution grid layer is designated for feeder load balancing, voltage quality control, and peak shaving and valley filling; and the microgrid layer is designated for local renewable energy consumption, islanded operation, and demand response. Establish hierarchical transmission paths between grid cells, feeders, substations, and main grid nodes.

[0028] As a preferred embodiment of the above, after the distribution network is divided into grids, a multi-level collaborative operation architecture covering the main grid layer, distribution network layer and microgrid layer is constructed based on the physical topology and operational boundary conditions of the power grid. The main grid layer is used to undertake energy market transactions and system-level frequency regulation and peak shaving demands, the distribution network layer is used for feeder load balancing, voltage quality control and peak shaving and valley filling, and the microgrid layer is used for local renewable energy consumption, islanded operation and demand response, so as to realize collaborative modeling of different levels of power grid under different time scales and functional positioning. Furthermore, in the multi-level power grid collaborative architecture, grid units are introduced as the basic units for refined sensing and control of the distribution network. Each grid unit corresponds to one or more transformer substations and has relatively independent load monitoring and control capabilities. At the same time, combined with power grid operation data, a mapping relationship is established between grid units and feeders, substations and main grid nodes, forming a hierarchical transmission path of "grid-feeder-substation-main grid node". This allows the main grid scheduling requirements, distribution network local constraints and microgrid autonomous operation requirements to be transmitted to the corresponding grid units level by level. Preferably, the multi-level power grid collaborative architecture is not limited to a three-level structure of main grid, distribution network, and microgrid. In practical applications, it can be simplified to a two-level structure or expanded to more levels depending on the power grid scale and operating scenario. Under different levels of architecture, the mapping relationship between grid units and upper-level distribution facilities can be used to realize the effective transmission of electric vehicle charging load prediction results, regulation potential boundaries, and regulation commands between different levels, thereby providing a unified architectural foundation for subsequent load forecasting, potential assessment, and coordinated regulation.

[0029] Furthermore, generating user behavior profiles corresponding to grid cells includes: Based on historical charging records and travel data analysis, the travel chain patterns of electric vehicle users within the grid unit are analyzed. Acquire information on electric vehicle users’ travel time distribution, charging preferences, State of Charge (SOC) trigger threshold, and target charging SOC level; Electric vehicle users are differentiated by clustering algorithms, and user behavior profiles are generated for each grid cell by combining the functional attributes of the grid cells.

[0030] As a preferred embodiment of the above, after completing the grid cell division and multi-source data preprocessing, based on historical charging records and user travel data, behavioral features are extracted and profiles are modeled for electric vehicle users within the grid cells. The user's historical travel trajectory, charging records and historical response behavior are analyzed to identify typical travel paths and stay duration distributions, so as to characterize the user's travel chain pattern. Furthermore, key behavioral characteristics of electric vehicle users are extracted from historical charging records and user travel data. These key behavioral characteristics include travel time distribution, charging preferences, frequently used charging periods, usual residence areas, state of charge trigger thresholds, target charging state of charge levels, and historical response characteristics. Among these, charging preferences can be reflected in information such as fast charging or slow charging, peak or off-peak hours, and home charging stations or public charging stations. Response characteristics can be reflected in the user's historical participation and fulfillment of obligations under different control signals or incentive conditions. Preferably, cluster analysis can be used to differentiate electric vehicle users into groups to identify several typical behavioral patterns. For example, user categories such as commuter-oriented, independent travel, and nighttime charging can be identified, and corresponding user behavior category labels can be formed. Subsequently, combined with the functional attributes of the grid unit to which the electric vehicle user belongs, the user behavior category labels are associated with the regional attributes to construct a three-dimensional profile database of "user-region-behavior" to generate user behavior profiles corresponding to the grid unit. The above methods enable a refined characterization of electric vehicle user behavior at the grid cell scale, providing a foundation for subsequent electric vehicle charging load forecasting, generalized energy storage equivalent modeling, and regulation potential assessment based on user behavior profiles, thereby improving the pertinence and accuracy of subsequent forecast results and regulation decisions.

[0031] Furthermore, predicting the electric vehicle charging load of each grid cell during the target time period includes: Establish a travel transfer probability model for electric vehicle users arriving at grid cells in the future; By integrating user behavior profiles and grid regulation signals corresponding to a multi-level grid collaborative architecture, the potential charging demand of each grid unit is generated. The predicted electric vehicle charging load for each grid cell within the target time period is output using a long short-term memory network or a Transformer model.

[0032] As a preferred embodiment of the above, after completing the construction of the user behavior profile, the electric vehicle charging load prediction model for grid units is established by combining the control requirements of different levels of the power grid in the multi-level power grid collaborative architecture and the power grid operation data, so as to make spatiotemporal predictions of the charging load of each grid unit within the target time period. Furthermore, based on historical travel trajectories and charging records, a travel transfer probability model is established for electric vehicle users to reach different grid units in the future, in order to predict the target grid units that users may reach in the future. On this basis, the travel chain characteristics, charging preferences and response characteristics in the user behavior profile are integrated to deduce the potential charging demand of each grid unit in the target period, thereby forming a predictive basis for mapping travel behavior to charging demand. Furthermore, based on potential charging demand, grid regulation signals corresponding to a multi-level grid collaborative architecture are introduced as external incentive variables. The main grid layer provides electricity price signals and peak-shaving demand signals, the distribution network layer provides load factor and voltage constraint information, and the microgrid layer provides local renewable energy output forecasts and energy storage status information. By coupling potential charging demand with grid regulation signals, a "travel-charging-grid" correlation prediction mechanism is constructed to characterize the charging load change patterns under different grid constraints and incentive conditions. Preferably, a long short-term memory network, a Transformer model, or other models suitable for time-series load forecasting can be used to predict the electric vehicle charging load of grid cells within a target time period and output the corresponding time-series load forecast results. The forecast results can characterize the distribution of electric vehicle charging load in each grid cell within a preset time range in the future and can support multi-scenario forecasting under normal operation scenarios, extreme weather scenarios, and special protection scenarios. The above methods enable refined prediction of electric vehicle charging load at the grid cell scale, allowing the prediction results to simultaneously reflect differences in user behavior, regional spatial differences, and multi-level grid coordination constraints, providing a basic input for subsequent generalized energy storage equivalent modeling, regulation potential assessment, and generation of differentiated control strategies.

[0033] Furthermore, constructing a generalized energy storage equivalent model and determining the regulation potential boundary of each grid cell includes: Based on the current state of charge, maximum charging power, and expected off-grid time of each electric vehicle in the grid cell, a set of available charging and discharging capabilities is constructed. The public charging stations within the grid unit are considered as aggregated generalized energy storage units; The Minkowski summation method is used to vector-superimpose the adjustable power ranges of N electric vehicles within a grid cell over the next T time periods to obtain the total adjustable power envelope.

[0034] As a preferred embodiment of the above, after obtaining the electric vehicle charging load prediction results of each grid unit in the target time period, the electric vehicle group in each grid unit is regarded as an aggregateable and flexibly adjustable resource. Combined with user behavior profiles and the operation constraints corresponding to the multi-level power grid collaborative architecture, the adjustable power of the electric vehicle group in the grid unit is aggregated and modeled to construct a generalized energy storage equivalent model corresponding to the grid unit. Furthermore, the public charging stations within each grid unit can be regarded as aggregated generalized energy storage units. For electric vehicles connected to the public charging stations, their available charging and discharging capabilities can be constructed based on parameters such as the current state of charge, maximum charging power, and expected off-grid time. The adjustable power range of each individual electric vehicle can be regarded as an aggregateable object to characterize the adjustable capability of the electric vehicle group within the grid unit during the target time period. Preferably, the Minkowski summation method can be used to vector-superimpose the adjustable power ranges of multiple individual electric vehicles within a grid cell to obtain the total adjustable power envelope of the public charging station in future time periods: Using the Minkowski summation method, the feasible power ranges of N individual EVs are vector-superimposed to obtain the total adjustable power envelope of the charging station in the future T time periods. ; in, and These represent the minimum and maximum adjustable power of the i-th EV at time t, respectively. In other implementations, Monte Carlo simulation, convolution operations, or other feasible domain aggregation and boundary calculation methods can also be used to achieve power aggregation and boundary calculation. Through the above aggregation modeling, the upper and lower limits of adjustable power and the equivalent energy storage capacity corresponding to the grid cell can be obtained, thereby constructing a generalized energy storage equivalent model. Furthermore, considering the dynamic access and exit characteristics of electric vehicles, the generalized energy storage equivalent model is updated in real time, so that the adjustable power aggregation result can change dynamically with the vehicle access status, and the total adjustable power envelope, adjustable power upper and lower limits, and energy storage equivalent capacity of the grid unit in the target time period are output. On this basis, combined with the operation requirements corresponding to the multi-level grid collaborative architecture, the adjustment potential boundary of the grid unit can be further formed, providing a basis for the subsequent assessment of the adjustable potential of electric vehicles and the generation of adjustment strategies.

[0035] Furthermore, determining the adjustment potential boundary for each grid cell includes: Construct a regulation potential indicator system, which should include at least transferable power, adjustable power, response speed, and reliability coefficient. The mechanistic constraint boundary of the grid cell is determined based on battery charging and discharging characteristics, vehicle range constraints, and users' minimum travel needs. A user response rate prediction model is trained based on historical response data, and the adjustment potential boundary of the grid cell is determined by combining the mechanistic constraint boundary.

[0036] As a preferred embodiment of the above, after completing the adjustable power aggregation of the electric vehicle group within the grid cell and constructing the generalized energy storage equivalent model, the actual regulation capability of the electric vehicle group within the grid cell is further evaluated to form the regulation potential boundary corresponding to the grid cell; wherein, the regulation potential boundary is used to characterize the range of actual regulation capability that the electric vehicle group within the grid cell can provide to the grid under the conditions of meeting user travel needs and battery operation constraints. Furthermore, a regulation potential index system can be constructed to quantitatively characterize the regulation capability of electric vehicle groups within a grid unit. The regulation potential index system includes at least transferable energy, adjustable power, response speed, and reliability coefficient. Among them, transferable energy is used to characterize the charging energy that can be transferred during off-peak hours, adjustable power is used to characterize the power regulation capability that supports peak shaving or valley filling, response speed is used to characterize the time delay from receiving the regulation command to executing the regulation, and reliability coefficient is used to characterize the user's fulfillment probability and response stability. Furthermore, the mechanistic constraint boundary of the grid unit is determined based on battery charging and discharging characteristics, vehicle range constraints, and users' minimum travel needs. Specifically, based on parameters such as electric vehicle battery capacity, state of charge, charging and discharging power limits, and expected off-grid time, combined with vehicle range requirements and users' minimum travel needs, the adjustable boundary of a single electric vehicle within the target time period is determined, and the mechanistic constraint boundary of the electric vehicle group within the grid unit is formed accordingly to ensure that the adjustment process does not affect users' normal vehicle usage needs. Preferably, a user response rate prediction model is trained based on historical response data to characterize the user's actual response capability to adjustment commands under different stimulus conditions. Subsequently, the output of the user response rate prediction model is fused with the mechanistic constraint boundary, and the adjustment potential boundary of the grid cell is determined by combining the mechanistic model and the data-driven model. This allows the adjustment potential boundary to not only reflect battery physical constraints and travel demand constraints, but also reflect the differences and uncertainties in user response behavior, thereby improving the authenticity and usability of the adjustment potential assessment results. The above methods enable dynamic assessment of the adjustment potential of electric vehicle groups at the grid cell scale, and provide a basis for subsequent generation of differentiated guidance strategies, issuance of adjustment instructions, and rolling corrections.

[0037] Furthermore, the electric vehicle charging load and adjustment potential boundary are rolled over based on real-time vehicle access status and historical response results, including: Match the regulation needs of the main grid layer, distribution network layer, and microgrid layer according to the regulation potential boundary; Generate a differentiated guidance strategy that includes at least one of electricity price incentives, credit points, and priority service rights, and output charging load adjustment instructions to the corresponding grid cells; Based on the previous forecast, the electric vehicle charging load and adjustment potential boundary are updated according to the latest vehicle access status at a preset rolling cycle of 15 minutes.

[0038] As a preferred embodiment of the above, after obtaining the adjustment potential boundary of each grid unit, the adjustment capability of the electric vehicle group within the grid unit is classified and matched according to the control requirements of different levels of the grid in the multi-level grid cooperative architecture, and a corresponding differentiated guidance strategy is generated accordingly. Furthermore, in the hierarchical matching process, the regulation requirements of the main grid layer may include system-level peak shaving, frequency regulation, and energy optimization requirements; the regulation requirements of the distribution grid layer may include feeder load balancing, voltage constraint control, and peak shaving and valley filling requirements; and the regulation requirements of the microgrid layer may include local renewable energy consumption, energy storage coordination, and local autonomous operation requirements. Based on the regulation objectives of different levels, the regulation potential boundary of the grid unit is matched with the regulation requirements of the corresponding level to determine the regulation direction, regulation magnitude, and response priority of each grid unit in the target time period. Preferably, the differentiated guidance strategy may include at least one of electricity price incentives, credit points, and priority service rights. The electricity price incentives are used to guide users to adjust their charging time through price signals, the credit points are used to increase users' enthusiasm for adjustment, and the priority service rights are used to increase the priority of responding users in subsequent charging services. After generating the differentiated guidance strategy, a charging load adjustment instruction is output to the corresponding grid unit to guide the electric vehicle group in the grid unit to participate in orderly charging, peak shaving and valley filling, demand response, or new energy consumption. Furthermore, to improve the adaptability of the control process to changes in actual operating conditions, the electric vehicle charging load and regulation potential boundary are rolled over based on real-time vehicle access status and historical response results. Specifically, based on the day-ahead forecast results and initial regulation potential boundary, and combined with the latest vehicle access status, actual user response behavior, and real-time grid operating status, the electric vehicle charging load and regulation potential boundary are periodically updated to achieve a shift from static planning to dynamic tracking. Preferably, the rolling correction can be performed according to a preset cycle to continuously update the adjustability and control strategy of each grid unit. Through the above methods, differentiated guidance and coordinated control under the different levels of power grid regulation needs can be achieved based on the regulation potential boundary. A closed-loop operation mechanism of "sensing-analysis-decision-feedback" can be formed through rolling correction, thereby improving the pertinence of regulation strategies, the accuracy of load regulation, and the reliability of electric vehicle groups participating in vehicle-grid interaction.

[0039] Example 2; Based on the same inventive concept as the load potential assessment method for a main-supplier-micro-cooperative electric vehicle in the foregoing embodiments, the present invention also provides a load potential assessment system for a main-supplier-micro-cooperative electric vehicle, the system comprising: The data acquisition module acquires the power grid topology information, charging behavior data, user travel data, power grid operation data, and regional functional attribute data of the target area. Based on the power grid topology information and regional functional attribute data, the distribution network is divided into grids to form dynamically adjustable grid units. The mapping construction module constructs a multi-level power grid collaborative architecture that integrates the main grid layer, distribution network layer, and microgrid layer based on grid cells and power grid operation data, and establishes the mapping relationship between grid cells and upper-level distribution facilities. The profile generation module extracts the travel characteristics, charging preferences, and response characteristics of electric vehicle users within the grid unit based on charging behavior data and user travel data, in order to generate user behavior profiles corresponding to the grid unit. The load forecasting module predicts the electric vehicle charging load of each grid unit within the target time period based on user behavior profiles, multi-level power grid collaborative architecture, and power grid operation data. The boundary adjustment module aggregates the adjustable power of electric vehicle groups within a grid unit based on electric vehicle charging load, user behavior profiles, and a multi-level power grid collaborative architecture, in order to construct a generalized energy storage equivalent model and determine the adjustment potential boundary of each grid unit. The instruction regulation module generates differentiated guidance strategies based on the regulation potential boundary and the regulation requirements corresponding to the multi-level power grid collaborative architecture, and outputs charging load regulation instructions to the corresponding grid units. It also performs rolling corrections on the electric vehicle charging load and regulation potential boundary based on the real-time vehicle access status and historical response results.

[0040] The system described above in this invention can effectively realize a method for assessing the load potential of a main-supplier-micro-cooperative electric vehicle, and the technical effects it can achieve are as described in the above embodiments, and will not be repeated here.

[0041] Example 3; Based on the same inventive concept as the method for assessing the load potential of a main-supplier-micro-cooperative electric vehicle in the foregoing embodiments, the present invention also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, can realize the method for assessing the load potential of a main-supplier-micro-cooperative electric vehicle.

[0042] Example 4; Figure 2 The following is a detailed explanation of the specific implementation process of this invention, using the overall architecture diagram and application scenarios in urban core areas as examples: 1. Distribution network grid division and data preparation Acquire geographic information system maps, distribution network topology data, and points of interest (POI) information for the target area. Based on the load distribution characteristics and functional zone attributes within the area, divide the distribution network into several distribution grid units (DGCs). Each DGC covers a certain geographical area and possesses relatively independent power supply characteristics. The system accesses historical electric vehicle charging data over a specific period, sourced from charging facility operation platforms, power distribution automation systems, meteorological service platforms, and distributed energy output monitoring systems. The raw data is cleaned, outlier or missing records are removed, time precision and units of measurement are standardized, and each charging event is matched to its corresponding Distributed Generation Control Center (DGC) number based on its geographical location, thus completing the spatial attribution labeling of the data.

[0043] 2. User behavior profile construction Based on historical charging records and travel data, cluster analysis was conducted on electric vehicle users within the region to identify several typical behavioral patterns, such as commuting-oriented, independent travel, and nighttime charging. Combining user charging preferences, frequently used time periods, residing areas, and historical response behaviors, a user behavior profile database was constructed to support subsequent personalized modeling.

[0044] 3. Load forecasting model training and deployment A time-series modeling method, including Long Short-Term Memory (LSTM) networks, is used to construct a predictive model for the correlation between travel, charging, and the power grid. Taking users' historical travel trajectories, charging behavior, and external adjustment signals as inputs, the model outputs the predicted electric vehicle charging load for each distributed grid collector (DGC) in future time periods. Furthermore, a generalized energy storage model was established for public charging stations, and the Minkowski summation method was used to calculate its adjustable power range during typical periods. The prediction results show that during the evening peak period on a typical workday, the charging load of the DGC in a certain commercial functional area is close to the transformer capacity limit, and there is a risk of local overload.

[0045] 4. Adjustable potential assessment and guidance strategy generation The system assesses the adjustable potential of the electric vehicle population within the DGC, comprehensively considering factors such as battery status, user travel demand, and historical response rates to estimate its transferable electricity and adjustable power capabilities. Simulations based on real data are also performed. Figure 3 As shown, the real-time adjustable potential is found to be much greater than the day-ahead forecast: for example, around 23:00 at night, the real-time available charging and discharging power exceeds the day-ahead average forecast by more than 40%. This phenomenon is due to the randomness of user behavior and the uncertainty of the actual vehicles connected to the grid. Therefore, this system introduces a rolling optimization mechanism: based on the day-ahead forecast, the adjustable boundary is updated every 15 minutes based on the latest vehicle access status, realizing the transformation from "static planning" to "dynamic tracking" and maximizing the exploitation of flexibility resources.

[0046] At the distribution network level, "Charging Station 1" is used as a representative node to illustrate whether it can accurately execute the main network scheduling instructions. Figure 4 As shown, the results indicate that the actual net power always operates within the dynamic boundary, and the terminal state of charge (SOC) approaches the target value, demonstrating that scheduling commands can be effectively transmitted. At the microgrid level, "Charging Station 3" is used as an independent microgrid unit to illustrate its robustness in highly random environments. Figure 5 As shown, the results demonstrate that even in the smallest unit, the model can still achieve a balance between peak shaving and valley filling and ensuring user demand.

[0047] 5. System closed-loop optimization Collect actual user response behavior data to update user profile tags and response probability models, thereby improving the accuracy of behavior prediction. The system regularly performs dynamic optimization of DGC boundary division, incentive parameter settings, and prediction model weights, forming a closed-loop operation mechanism of "perception-analysis-decision-feedback" to continuously improve the system's adaptive capabilities and control efficiency.

[0048] This implementation fully demonstrates the feasibility and effectiveness of the present invention in a real urban environment, and verifies that the technical path based on the combination of distribution network grid management, user behavior modeling and main distribution micro-coordinated control can achieve accurate prediction of electric vehicle charging load and efficient utilization of adjustable resources.

[0049] like Figure 6 The multi-source data-driven dynamic spatial grid partitioning method used in this invention is employed to construct the spatial basis unit for predicting charging demand for electric private vehicles. This method comprehensively considers the spatial distribution characteristics of user travel behavior and the attributes of urban functional areas, achieving scientific zoning of the study area and providing a high-precision geographic reference framework for subsequent charging load modeling. Excessive spatial scale can lead to a large number of vehicle origin-destination (OD) trajectories clustering within the same grid, resulting in invalid travel trajectory data; conversely, insufficient spatial scale can lead to an increase in the number of empty trajectory grids and increased computational difficulty in the model. Determining the functional attributes of the geographic grid is equally important for establishing the functional area grid model. Since different types of Points of Interest (POIs) exhibit significant differences in their influence on surrounding areas, further analysis and processing of the acquired POI data are necessary to obtain the functional attributes of the geographic grid.

[0050] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for assessing the load potential of a main-supplier micro-cooperative electric vehicle, characterized in that, The method includes: The system acquires power grid topology information, charging behavior data, user travel data, power grid operation data, and regional functional attribute data for the target area. Based on the power grid topology information and the regional functional attribute data, the distribution network is divided into grids to form dynamically adjustable grid units. Based on the grid units and the power grid operation data, a multi-level power grid collaborative architecture is constructed, which coordinates the main grid layer, distribution network layer, and microgrid layer, and a mapping relationship is established between the grid units and the upper-level distribution facilities. Based on the charging behavior data and the user travel data, the travel characteristics, charging preferences and response characteristics of electric vehicle users within the grid unit are extracted to generate a user behavior profile corresponding to the grid unit; Based on the user behavior profile, the multi-level power grid collaborative architecture, and the power grid operation data, predict the electric vehicle charging load of each grid unit during the target time period; Based on the electric vehicle charging load, the user behavior profile, and the multi-level power grid collaborative architecture, the adjustable power of the electric vehicle group within the grid unit is aggregated to construct a generalized energy storage equivalent model and determine the adjustment potential boundary of each grid unit. Based on the regulation potential boundary and the regulation requirements corresponding to the multi-level power grid collaborative architecture, a differentiated guidance strategy is generated, and a charging load regulation command is output to the corresponding grid unit. The electric vehicle charging load and the regulation potential boundary are then rolled over according to the real-time vehicle access status and historical response results.

2. The method for assessing the load potential of a main-distributor micro-cooperative electric vehicle according to claim 1, characterized in that, The distribution network is divided into grids, including: Acquire geographic information system data, power distribution network topology data, and point of interest (POI) data for the target area; The grid cell boundaries are determined based on the geographic information system data, the power distribution network topology data, the regional load density, the power supply radius, and the point of interest (POI) data. The boundaries of the grid cells are dynamically adjusted based on changes in load hotspots.

3. The method for assessing the load potential of a main-supplier-micro-cooperative electric vehicle according to claim 1, characterized in that, Establishing the mapping relationship between the grid cells and the upper-level power distribution facilities includes: A collaborative operation framework consisting of a main grid layer, a distribution grid layer, and a microgrid layer is constructed based on the aforementioned grid units. The main grid layer is designated for energy market trading and system-level frequency regulation and peak shaving, the distribution grid layer is designated for feeder load balancing, voltage quality control and peak shaving and valley filling, and the microgrid layer is designated for local renewable energy consumption, islanded operation and demand response. Establish hierarchical transmission paths between the grid cells, feeders, substations, and main grid nodes.

4. The method for assessing the load potential of a main-supplier-micro-cooperative electric vehicle according to claim 1, characterized in that, Generating user behavior profiles corresponding to the grid cells includes: Based on historical charging records and travel data analysis, the travel chain patterns of electric vehicle users within the grid unit are analyzed. The travel time distribution, charging preferences, state of charge (SOC) trigger threshold, and target charging SOC level of the electric vehicle users are obtained. The electric vehicle users are differentiated into groups using a clustering algorithm, and the user behavior profile corresponding to the grid unit is generated by combining the functional attributes of the grid unit.

5. The method for assessing the load potential of a main-supplier-micro-cooperative electric vehicle according to claim 1, characterized in that, Predicting the electric vehicle charging load of each grid cell during the target time period includes: Establish a travel transfer probability model for the electric vehicle users arriving at the grid cell in the future; The potential charging demand of each grid unit is generated by integrating the user behavior profile and the grid regulation signal corresponding to the multi-level power grid collaborative architecture. The predicted electric vehicle charging load for each grid cell within the target time period is output using a long short-term memory network or a Transformer model.

6. The method for assessing the load potential of a main-supplier micro-cooperative electric vehicle according to claim 5, characterized in that, Constructing a generalized energy storage equivalent model and determining the regulation potential boundary of each grid cell includes: Based on the current state of charge, maximum charging power, and expected off-grid time of each electric vehicle in the grid cell, a set of available charging and discharging capabilities is constructed. The public charging stations within the grid unit are considered as aggregated generalized energy storage units. The Minkowski summation method is used to vector-superimpose the adjustable power ranges of N electric vehicles within the grid cell over the next T time periods to obtain the total adjustable power envelope.

7. The method for assessing the load potential of a main-supplier micro-cooperative electric vehicle according to claim 6, characterized in that, Determining the adjustment potential boundary of each of the aforementioned grid cells includes: Construct a regulation potential index system, which includes at least transferable power, adjustable power, response speed, and reliability coefficient. The mechanistic constraint boundary of the grid cell is determined based on battery charging and discharging characteristics, vehicle range constraints, and users' minimum travel requirements. The user response rate prediction model is trained based on historical response data, and the adjustment potential boundary of the grid cell is determined by combining the mechanistic constraint boundary.

8. The method for assessing the load potential of a main-supplier-micro-cooperative electric vehicle according to claim 1, characterized in that, The electric vehicle charging load and the adjustment potential boundary are rolled over based on real-time vehicle access status and historical response results, including: The regulation requirements of the main grid layer, the distribution grid layer, and the microgrid layer are matched according to the regulation potential boundary; Generate a differentiated guidance strategy that includes at least one of electricity price incentives, credit points, and priority service rights, and output a charging load adjustment command to the corresponding grid cell; Based on the previous forecast, the electric vehicle charging load and the adjustment potential boundary are updated according to the latest vehicle access status at a preset rolling period of 15 minutes.

9. A load potential assessment system for a main-supplier-micro-cooperative electric vehicle, characterized in that, The system includes: The data acquisition module acquires the power grid topology information, charging behavior data, user travel data, power grid operation data, and regional functional attribute data of the target area. Based on the power grid topology information and regional functional attribute data, the distribution network is divided into grids to form dynamically adjustable grid units. The mapping construction module constructs a multi-level power grid collaborative architecture that integrates the main grid layer, distribution network layer, and microgrid layer based on grid cells and power grid operation data, and establishes the mapping relationship between grid cells and upper-level distribution facilities. The profile generation module extracts the travel characteristics, charging preferences, and response characteristics of electric vehicle users within the grid unit based on charging behavior data and user travel data, in order to generate user behavior profiles corresponding to the grid unit. The load forecasting module predicts the electric vehicle charging load of each grid unit within the target time period based on user behavior profiles, multi-level power grid collaborative architecture, and power grid operation data. The boundary adjustment module aggregates the adjustable power of electric vehicle groups within a grid unit based on electric vehicle charging load, user behavior profiles, and a multi-level power grid collaborative architecture, in order to construct a generalized energy storage equivalent model and determine the adjustment potential boundary of each grid unit. The instruction regulation module generates differentiated guidance strategies based on the regulation potential boundary and the regulation requirements corresponding to the multi-level power grid collaborative architecture, and outputs charging load regulation instructions to the corresponding grid units. It also performs rolling corrections on the electric vehicle charging load and regulation potential boundary based on the real-time vehicle access status and historical response results.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which includes program instructions that, when executed by a processor, implement the method for assessing the load potential of a primary-secondary-micro cooperative electric vehicle as described in any one of claims 1-8.

Images