New energy vehicle user schedulable potential grading evaluation method and system based on multi-modal data fusion

Abstract

This paper presents a hierarchical assessment method and system for the dispatchable potential of new energy vehicle users based on multimodal data fusion. Belonging to the field of power system technology, the key technical points include: collecting primary energy meter data from smart charging piles for new energy vehicles over a certain period; preprocessing the data and calculating electricity consumption data; converting time-series data into multidimensional features through feature engineering; extracting common factors using factor analysis and clustering to identify typical consumption patterns of electric vehicle users; constructing identification tags for electric vehicle user demand response behavior; analyzing the mechanism of electric vehicle user demand response behavior; analyzing the effects of differentiated interventions on demand response and heterogeneous groups; and analyzing the overall response effect of differentiated pricing strategies.

Description

Technical Field

[0001] This invention belongs to the field of power system technology, specifically relating to a method and system for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion. Background Technology

[0002] While renewable energy generation accounts for a high proportion, its stability is poor, leading to a significant supply-demand imbalance in winter. Southern regions face considerable regulatory pressure due to extreme weather and the widespread adoption of electric vehicles. To improve demand response for electric vehicle users, this study integrates policy data, questionnaires, and meteorological data to analyze user behavior characteristics and design incentive programs. Summary of the Invention

[0003] To address the aforementioned problems, this application provides a method for hierarchical assessment of the dispatchable potential of new energy vehicle users based on multimodal data fusion, comprising the following embodiments:

[0004] Data from the electricity meter of a smart charging station for new energy vehicles is collected once at regular intervals.

[0005] Data preprocessing and calculation of electricity consumption data;

[0006] Transform time series data into multidimensional features through feature engineering;

[0007] Factor analysis and clustering were used to extract common factors and identify typical consumption patterns of electric vehicle users.

[0008] Constructing identification tags for electric vehicle user demand response behavior;

[0009] Analysis of the mechanism of electric vehicle user demand response behavior;

[0010] Analysis of the effects of differentiated interventions in demand response and heterogeneous groups;

[0011] Analysis of the overall response effect of differentiated pricing strategies.

[0012] According to some embodiments of this application, a method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion is provided, wherein the step of data preprocessing includes removing users with missing monthly data.

[0013] Among them, the step transforms time series data into multi-dimensional features through feature engineering, including decomposing complex electricity consumption waveforms into intrinsic mode functions through time series data feature mining and empirical mode decomposition, and capturing local features at each time scale.

[0014] The steps include using factor analysis clustering to extract common factors and identify typical consumption patterns of electric vehicle users. This includes constructing key features based on monthly electricity consumption data, calculating factor scores through factor analysis, and labeling user electricity consumption patterns using the K-means clustering algorithm to achieve typical pattern recognition.

[0015] According to some embodiments of this application, a hierarchical assessment method for the dispatchable potential of new energy vehicle users based on multimodal data fusion is provided, wherein key features include electricity consumption time periods, annual average, variance, skewness, kurtosis, frequency, period, and inflection point information of each trend.

[0016] According to some embodiments of this application, a hierarchical assessment method for the dispatchable potential of new energy vehicle users based on multimodal data fusion is proposed, wherein the step of constructing identification tags for electric vehicle user demand response behavior includes exploring the inherent patterns of user behavior through intelligent demand response questionnaire data and cluster learning, and automatically mining and constructing a tag library.

[0017] According to some embodiments of the present application, a hierarchical assessment method for the dispatchable potential of new energy vehicle users based on multimodal data fusion includes a step of analyzing the demand response behavior mechanism of electric vehicle users, including...

[0018] Demand response activity participation intention analysis: Using a binary logit model, key factors influencing electric vehicle users' willingness to participate in energy-saving activities are identified from user personal attributes, hardware attributes, necessary travel frequency, and charging mode. The model is constructed as follows:

[0019]

[0020] Demand response activity participation analysis: Based on the actual electricity savings of users, a multiple regression model is constructed to measure the impact of the above factors on the activity participation level. The multiple regression model is as follows:

[0021] Resp i =γ0+γ i Att i +γ2X i .

[0022] According to some embodiments of the present application, a hierarchical assessment method for the dispatchable potential of new energy vehicle users based on multimodal data fusion includes a step of analyzing the demand response behavior mechanism of electric vehicle users, including...

[0023] To identify the overall response effect of cost incentives on invited users, the following model is constructed:

[0024] E_usage it =γ0+γ1(T) i ×Att t )+γ3Ti +γ4Att t +ε it

[0025] By controlling the effects on different users at different comparable levels, an extended effect evaluation model is constructed as follows:

[0026] E_usage it =γ0+γ1(T) i ×Att t )+γ3T i +γ4Att t +γ5X it +μ i +ε it .

[0027] Based on the multimodal data fusion-based hierarchical evaluation method for the dispatchable potential of new energy vehicle users according to some embodiments of this application, and the response effect analysis of electric vehicle user groups based on information incentives, the model is constructed as follows:

[0028] E_usage it =γ0+γ1(T) i ×Inv t )+γ2T i +γ3Inv i +γ4X it +μ i +ε it .

[0029] According to some embodiments of the present application, a method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion includes a step of analyzing the overall response effect of differentiated pricing strategies.

[0030] The impact of economic subsidy strategies on charging and discharging behavior was investigated, and two subsidy schemes were designed:

[0031] Incremental strategy: Initial subsidies are low, then gradually increased;

[0032] User acquisition strategy: Initial subsidies are higher, then gradually decrease;

[0033] The dual fixed effects regression model is constructed as follows:

[0034] Save_ele it =β0+β1Pri it +β2Base it +α i +T t +ε it .

[0035] This application also provides an electronic device, which includes: one or more processors, a memory, and one or more programs; wherein the one or more programs are stored in the memory, and the one or more programs include instructions that, when executed by the electronic device, cause the electronic device to perform a first aspect and any possible technical solution of the first aspect.

[0036] This application also provides a computer-readable storage medium, which includes a computer program that, when run on an electronic device, causes the electronic device to perform the first aspect and any possible technical solution of the first aspect.

[0037] Beneficial effects:

[0038] Multi-source data fusion and feature mining: Construct a high-response multi-source database, integrate electricity consumption data (AMI) and questionnaire data (consumer psychology, family structure), and use clustering algorithms, empirical mode decomposition (EMD) and neural network models to establish a user behavior tag library with optimized prediction accuracy.

[0039] Causal inference and hybrid research paradigm: Based on randomized controlled trials (RCTs) to quantify intervention effects, a research method integrating model-driven (power system simulation) and data-driven (user response data) approaches is established to meet the dual needs of correlation analysis and causal inference in the "source-load interaction" scenario.

[0040] Strategy self-learning optimization: Develop a user historical response performance evaluation model to achieve dynamic optimization of demand response strategies and improve invitation accuracy and execution efficiency. Attached Figure Description

[0041] Figure 1 Feature self-learning model. Detailed Implementation

[0042] The embodiments of this application are described in detail below with reference to the accompanying drawings, examples of which are shown in the drawings. This application provides a method and an electronic device. The method and device are based on the same technical concept. Since the principles by which the method and device solve the problem are similar, the implementations of the device and method can be referred to each other, and repeated details will not be repeated. 1. Overview of the Invention

[0044] Electricity demand response, as an important tool for demand-side management, is widely used globally; however, research on integrated energy services and demand response for electric vehicles is still in its early stages. This invention combines user questionnaire data and actual charging behavior to conduct an in-depth analysis from a holistic and heterogeneous perspective, providing a new theoretical perspective for this field.

[0045] This invention can provide a policy basis for promoting the marketization of V2G, help power grid companies and integrated energy service providers optimize load management and dispatch, and help electric vehicle users reduce energy costs and increase peak-shaving benefits by participating in demand response.

[0046] Electric vehicle demand response (EVD) is a short-term load management mechanism that guides users to adjust their charging and discharging behavior, helping them optimize electricity consumption and reduce costs, while providing the power grid with flexible adjustment capabilities and promoting the construction of a low-carbon energy system. This mechanism is of great significance for the safe and stable operation of the power grid, economic dispatch, and alleviating peak electricity demand.

[0047] 2. Research and Implementation Plan: This invention is guided by the electricity demand of electric vehicle users. Based on smart charging pile data, it integrates survey data on user economic status, demographic characteristics, and consumption habits to construct a user electricity consumption characteristic database and a demand response behavior labeling system. In pilot projects (such as at the provincial level), an invitation-based demand response system is implemented through economic incentives and behavioral guidance, and regional potential and costs are analyzed.

[0048] Under the "Power Grid Company-User" interaction model: 1) The power grid company pushes incentive strategies through App / SMS / Official Account; 2) Users participate in the response through off-peak charging / V2G discharge; 3) They receive electricity fee red envelopes / points rewards; 4) A sustainable ecosystem of "user benefit-power grid peak shaving" is formed.

[0049] 2.1 Survey Design:

[0050] (1) Preliminary preparation: Obtain professional insights through literature review and improve the research plan.

[0051] (2) Design questionnaires: Design online and offline questionnaires according to the research objectives to ensure the validity of the data.

[0052] (3) Preliminary survey: The reliability and validity were verified through questionnaire collection and field visits, and a preliminary analysis was conducted.

[0053] (4) Large-scale survey: Combine customer service center communication, community visits and online questionnaires to expand the sample size.

[0054] (5) Questionnaire follow-up: Screen the data and conduct telephone follow-ups and on-site verification to ensure the quality of the survey.

[0055] 2.2 Research and Practice Plan and Expected Results

[0056] 2.2.1 Questionnaire Survey Implementation: Based on the Irish CBT framework, this invention designed a 47-question questionnaire to investigate the demand response behavior of electric vehicle users, focusing on issues such as owner information, vehicle usage behavior, frequency, and electricity usage concepts, revealing the key factors influencing response willingness.

[0057] In the summer of 2024, the project team collected more than 40,000 questionnaires in a certain region through an online platform and used 100% telephone follow-up to ensure data quality. In the end, 13,000 valid samples were obtained, and the largest and most comprehensive smart energy service database for electric vehicle users was established.

[0058] 2.2.2 Related Platform Construction: An electric vehicle demand response platform will be built based on the power supply service command platform to realize functions such as demand response, smart payment, and charging / discharging analysis.

[0059] 2.2.3 Expected Effects of Electric Vehicle Demand Response: It is estimated that 150,000 electric vehicle users will be invited through the demand response activities, of which 50,000 users will participate in peak-shifting electricity consumption and peak-discharge using V2G technology. It is estimated that each household can reduce its electricity consumption by 7KW and discharge 7KW in reverse. The 50,000 users can reduce peak load by a total of 350MW and provide 350MW of discharge support.

[0060] 2.2.4 Response Period Selection: The peak daily load of a power grid in a certain region during summer is positively correlated with temperature, while the peak daily load of a power grid in a certain region during winter is negatively correlated with temperature. To ensure peak shaving effect at the user end, dates with higher and lower temperatures are selected as demand response days, and it is ensured that the temperature on the response day in both locations is consistent with the baseline day during the assessment.

[0061] 3. Survey Content and Practical Results: 3.1 Feature Engineering and Tag Library Construction Based on Electric Vehicle Electricity Consumption Data and Micro-survey Data

[0062] 3.1.1 Data Preprocessing: Smart charging piles were installed in the pilot area, and electricity meter data was collected every 15 minutes (a total of 96 data points) to provide a basis for analyzing the demand response behavior of electric vehicle users. For ease of processing, the data for a single user per day was stored as a vector in time series format. For missing values ​​caused by signal interference or equipment failure, linear imputation was performed using the mean of the preceding and following data. Simultaneously, the load curves of different users were normalized (mapped to 0-1) to ensure accurate feature identification.

[0063] 3.1.2 Architecture Analysis: First, the massive amount of collected electricity consumption data is preprocessed; then, feature engineering is used to convert the time series data into multi-dimensional features to reduce noise and uncover potential trends; finally, factor analysis is used for clustering to extract common factors and identify typical consumption patterns of electric vehicle users.

[0064] 3.1.3 Feature Engineering Construction: Effective features are extracted from the original data through feature construction, selection, and dimensionality reduction to improve clustering performance. Considering that data interpolation may introduce bias and that a small number of random missing data entries do not affect overall representativeness, users with missing monthly data were removed during the preprocessing stage, retaining only samples with monthly electricity consumption records. Time-series data feature mining (Python tsfresh module) and Empirical Mode Decomposition (EMD) are used to decompose complex electricity consumption waveforms into intrinsic mode functions, capturing local features at each time scale. Finally, 42 key features (such as electricity consumption time period, annual mean, variance, skewness, kurtosis, frequency, period, and inflection point information of each trend) are constructed based on monthly electricity consumption data. Factor scores are calculated through factor analysis, and K-means clustering algorithm is used to label user electricity consumption patterns, achieving typical pattern recognition.

[0065] 3.2 Identification Tags for Electric Vehicle User Demand Response Behavior and Potential Customer Matching Model

[0066] 3.2.1 User Profile and Tag Construction: Through intelligent demand response questionnaire data and cluster learning, the inherent patterns of user behavior are explored, and a tag library is automatically mined and constructed (see...). Figure 1 The data sources include 24 months of monthly electricity consumption data from electric vehicle users and 7 days of charging station data questionnaires before and after the response. After cleaning, transformation, standardization, and missing data processing, the data was divided into training and test sets in a 7:3 ratio. Based on the questionnaire design, multi-dimensional information such as urban and rural users' vehicle usage patterns, frequency of outings, and temperature sensitivity was extracted. Label coding was used to reconstruct the features, forming an adaptive label library.

[0067] 3.3 Identification of Key Factors and Effect Analysis of Demand Response Behavior

[0068] 3.3.1 Analysis of the Mechanism of Electric Vehicle User Demand Response Behavior: Based on the pilot practice of electric vehicle user demand response in a certain region, and combined with the electric vehicle user questionnaire data in the specimen database, this study explores the key influencing factors of electric vehicle user families' willingness and degree of participation in demand response activities from the perspective of individual electric vehicle user attributes, and analyzes the mechanism of electric vehicle user demand response behavior.

[0069] (1) Analysis of willingness to participate in demand response activities: First, based on the questionnaire data, a dummy variable "participation" was defined. Using a binary logit model, key factors influencing the willingness of electric vehicle users to participate in energy-saving activities were identified from aspects such as user personal attributes, hardware attributes, necessary travel frequency, and charging mode. The model is then constructed as follows:

[0070]

[0071] (2) Analysis of the degree of participation in demand response activities

[0072] Secondly, based on the actual electricity savings of users, a multiple regression model is constructed to measure the impact of the above factors on the level of participation in the activity. The multiple regression model is as follows:

[0073] Resp i =γ0+γ i Att i +γ2X i (3.2)

[0074] 3.3.2 Effects of Differentiated Intervention in Demand Response and Analysis of Heterogeneous Groups: Based on peak summer and peak winter electricity consumption practices in a specific region, charging pile data was collected every 15 minutes and combined with an experimental sample library to scientifically evaluate the response effects of invited users, quantifying the impact of "fee" compensation and "information" stimulation. A difference-in-difference model, based on a counterfactual framework, was used to assess the changes in electricity consumption for different types of users regardless of whether demand response occurred.

[0075] (1) Analysis of the response effects of different intervention strategies based on cost incentives on electric vehicle user groups

[0076] To identify the overall response effect of the cost incentive on users who participate in the invitation, the following model is constructed:

[0077] E_usage it =γ0+γ1(T) i ×Att t )+γ3T i +γ4Att t +ε it (3.3)

[0078] To further control the comparability of samples, this invention incorporates fixed effects into the model. The models are clustered into three levels: city level, urban-rural level, and different neighborhoods, thus controlling the effects on different users at different comparable levels. The extended effect evaluation model is constructed as follows:

[0079] E_usage it =γ0+γ1(T) i ×Att t )+γ3T i +γ4Att t +γ5X it +μ i +ε it (3.4)

[0080] After excluding natural electricity consumption, electricity consumption by participating users decreased significantly, with an estimated average reduction of approximately 0.5 kWh compared to the baseline date. Including monthly average electricity consumption and telephone advertising variables, the price compensation is expected to effectively reduce electricity consumption on the response date.

[0081] (2) Analysis of the response effect of electric vehicle users based on information incentives: By comparing users who only receive information promotion (without financial compensation) with users who have not participated in the activities at all, a sample analysis was conducted to analyze the difference in the response and electricity consumption behavior of users who were "only" stimulated by information and users who were not stimulated by any information, and a corresponding model was constructed to evaluate the effect of information promotion.

[0082] To identify the effect of information dissemination on the response to the needs of electric vehicle users, the following model is constructed:

[0083] E_usage it =γ0+γ1(T) i ×Inv t )+γ2T i +γ3Inv i +γ4X it +μ i +ε it (3.6)

[0084] The model infers that invited users did not respond to the SMS message alone. Electric vehicle users who received only the SMS message without any compensation incentive did not show significant differences in their electricity consumption behavior compared to those who received no SMS message at all; therefore, compensation is indispensable in demand response.

[0085] 3.3.3 Analysis of the Overall Response Effect of Differentiated Pricing Strategy

[0086] (1) Comparison of the response effects of differentiated pricing strategies: To further differentiate the impact of economic subsidy strategies on charging and discharging behavior, two subsidy schemes were designed:

[0087] Incremental strategy: The initial subsidy is low, and then it is gradually increased (the pilot price is 2 yuan / kWh, 5 yuan / kWh, and 8 yuan / kWh);

[0088] Customer acquisition strategy: Initial subsidies are higher, then gradually decrease (pilot pricing is 10 yuan / kWh, 8 yuan / kWh, and 5 yuan / kWh).

[0089] The dual fixed effects regression model is constructed as follows:

[0090] Save_ele it =β0+β1Pri it +β2Base it +α i +T t +ε it(3.7)

[0091] Therefore, it can be inferred that the subsidy price has a significant impact on energy savings in both the incremental strategy and the new customer acquisition strategy. Furthermore, the new customer acquisition strategy is more effective in incentivizing individual electric vehicle users.

[0092] Application scenarios

[0093] 1. Power grid service support: (1) Demand response package design: optimize parameters such as time-of-use pricing and V2G incentives; (2) Load forecasting and regulation: predict regional load curves based on the tag library; (3) Energy efficiency management: generate user-level electricity consumption analysis reports.

[0094] 2. Policy formulation support: (1) Electricity pricing policy: analyze user behavior elasticity; (2) Electric vehicle promotion: identify high-potential user groups; (3) Carbon metering management: establish "electricity-carbon" related accounts.

[0095] 3. Technology Expansion: Adaptable to new interaction modes such as V2G and V2X, supporting large-scale grid-connected management of electric vehicles. It is projected that the global electric vehicle fleet will reach 140 million to 240 million vehicles by 2030. This framework can enhance the power system's ability to coordinate and regulate resilient loads through a real-time updated tag library.

[0096] Based on the above embodiments, this application also provides a computer storage medium storing a computer program, which, when executed by a computer, causes the computer to perform the methods provided in the above embodiments.

[0097] The storage medium can be any available medium that a computer can access. For example, but not limited to, a computer-readable medium can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.

[0098] Based on the above embodiments, this application also provides a chip for reading a computer program stored in a memory to implement the method provided in the above embodiments.

[0099] Based on the above embodiments, this application provides a computer program product that implements the methods provided in the above embodiments when the computer program product is run on an electronic device.

[0100] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0101] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0102] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0103] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0104] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A hierarchical assessment method for the dispatchable potential of new energy vehicle users based on multimodal data fusion, characterized in that, include Data from the electricity meter of a smart charging station for new energy vehicles is collected once at regular intervals. Data preprocessing and calculation of electricity consumption data; Transform time series data into multidimensional features through feature engineering; Factor analysis and clustering were used to extract common factors and identify typical consumption patterns of electric vehicle users. Constructing identification tags for electric vehicle user demand response behavior; Analysis of the mechanism of electric vehicle user demand response behavior; Analysis of the effects of differentiated interventions in demand response and heterogeneous groups; Analysis of the overall response effect of differentiated pricing strategies.

2. The method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion according to claim 1, characterized in that, in, The steps involve data preprocessing, including removing users with missing month data; Among them, the step transforms time series data into multi-dimensional features through feature engineering, including decomposing complex electricity consumption waveforms into intrinsic mode functions through time series data feature mining and empirical mode decomposition, and capturing local features at each time scale. The steps include using factor analysis clustering to extract common factors and identify typical consumption patterns of electric vehicle users. This includes constructing key features based on monthly electricity consumption data, calculating factor scores through factor analysis, and labeling user electricity consumption patterns using the K-means clustering algorithm to achieve typical pattern recognition.

3. The method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion according to claim 1, characterized in that, in, Key features include electricity usage time periods, annual average, variance, skewness, kurtosis, frequency, period, and inflection point information for each trend.

4. The method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion according to claim 1, characterized in that, in, The steps involve constructing identification tags for electric vehicle user demand response behavior, including exploring the inherent patterns of user behavior through intelligent demand response questionnaire data and cluster learning, and automatically mining and constructing a tag library.

5. The method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion according to claim 1, characterized in that, in, The steps involve an analysis of the mechanism of electric vehicle user demand response behavior, including Demand response activity participation intention analysis: Using a binary logit model, key factors influencing electric vehicle users' willingness to participate in energy-saving activities are identified from user personal attributes, hardware attributes, necessary travel frequency, and charging mode. The model is constructed as follows: Demand response activity participation analysis: Based on the actual electricity savings of users, a multiple regression model is constructed to measure the impact of the above factors on the activity participation level. The multiple regression model is as follows: Reply i =γ0+γ i Att i +γ2X i 。 6. The method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion according to claim 1, characterized in that, in, The steps involve an analysis of the mechanism of electric vehicle user demand response behavior, including To identify the overall response effect of cost incentives on invited users, the following model is constructed: E_usage it =γ0+γ1(T i ×Att t )+γ3T i +γ4Att t +e it By controlling the effects on different users at different comparable levels, an extended effect evaluation model is constructed as follows: E_usage it =γ0+γ1(T i ×Att t )+γ3T i +γ4Att t +γ5X it +m i +e it 。 7. The method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion according to claim 1, characterized in that, Based on the analysis of the response effect of electric vehicle users with information incentives, the model is constructed as follows: E_usage it =γ0+γ1(T i ×Inv t )+γ2T i +γ3Inv i +γ4X it +m i +e it 。 8. The method for hierarchical evaluation of the dispatchable potential of new energy vehicle users based on multimodal data fusion according to claim 1, characterized in that, in, The steps involved in the overall response effect analysis of the differentiated pricing strategy include... The impact of economic subsidy strategies on charging and discharging behavior was investigated, and two subsidy schemes were designed: Incremental strategy: Initial subsidies are low, then gradually increased; User acquisition strategy: Initial subsidies are higher, then gradually decrease; The dual fixed effects regression model is constructed as follows: Save_ele it =β0+β1Pri it +β2Base it +a i +T t +e it 。 9. An electronic system, the system comprising: One or more processors, a memory, and one or more programs; wherein the one or more programs are stored in the memory, and the one or more programs include instructions that, when executed by the electronic device, cause the electronic device to perform the method of any one of claims 1-8.

10. A computer-readable storage medium comprising a computer program that, when executed on an electronic device, causes the electronic device to perform the method of any one of claims 1-8.

Images