Electric vehicle participation V2G potential calculation method based on user travel habits

Abstract

The invention discloses a user travel habit-based electric vehicle participation V2G potential calculation method, and particularly relates to the technical field of electric vehicle management, and the method comprises the following steps: precisely recognizing whether a user has a sign of using an electric vehicle to participate in V2G potential calculation through power supply intention investigation; for users with potential signs, travel habit data of the users are obtained, electric quantity potential features and time window potential features are extracted, electric quantity potential indexes and time window potential indexes are generated, the users are classified, low-potential users are eliminated, the optimization direction of middle-potential users is deduced through fuzzy reasoning, and personalized improvement suggestions are provided for the users; through accurate screening and optimization, efficient allocation of resources is realized, the application efficiency of the electric vehicle in the V2G mode is improved, the travel demand of a user is guaranteed, the load balance of a power grid and the utilization rate of renewable energy sources are optimized, green energy development is promoted, and the method has wide application value.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle management, and more specifically, to a method for calculating the potential of electric vehicles to participate in V2G based on user travel habits. Background Art

[0002] With the development of new energy technologies, electric vehicles (EVs) have gradually become one of the important means of transportation in modern society. In order to improve energy efficiency and promote the popularization of renewable energy, electric vehicles can not only be used as a means of transportation, but also as a distributed energy storage unit to participate in the operation and management of the power grid. This mode is called vehicle-to-grid (V2G). Through the V2G mode, electric vehicles feed their electricity back to the grid during non-use periods, providing support for load balancing, peak-valley regulation and optimal utilization of renewable energy.

[0003] However, the current promotion of the V2G model faces the following technical challenges: Complexity of user travel habits: There are significant differences in parking time, parking behavior regularity, and power usage requirements among different users, making it difficult to adapt a unified model. Resource waste problem: Failure to scientifically screen the user's participation potential leads to uneven resource allocation and reduces V2G operation efficiency. Therefore, a method for calculating the potential of electric vehicles to participate in V2G based on user travel habits is proposed here to accurately screen out users suitable for the V2G model and improve the user's potential to participate in V2G through optimization guidance. Summary of the invention

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A method for calculating the potential of electric vehicles to participate in V2G based on user travel habits, comprising the following steps:

[0006] Conduct a survey on users' willingness to supply electricity, then conduct a survey and analysis based on the survey results, and identify whether the users have signs that their electric vehicles can participate in the V2G potential calculation;

[0007] If the user has signs that the electric vehicle used by the user participates in the V2G potential calculation, the user's travel habits are sequentially subjected to information acquisition, feature extraction, and feature analysis operations, and then the user's travel habits are classified according to the results of the feature analysis operations, and the user is divided into a high potential travel type, a medium potential travel type, and a low potential travel type;

[0008] Users with low potential travel types are eliminated, and fuzzy reasoning is used to infer the improvement direction of users with medium potential travel types.

[0009] In a preferred embodiment, the power supply willingness survey includes an economic return interest survey, a green energy interest survey, and an incentive mechanism participation interest survey.

[0010] In a preferred embodiment, the investigation and analysis refers to:

[0011] The corresponding interest scores of users in the economic return interest survey, green energy interest survey, and incentive mechanism participation interest survey are obtained respectively, and then the average of the interest score values ​​corresponding to the three types of interests is calculated to obtain the average interest value, and the standard deviation of the interest score values ​​corresponding to the three types of interests is calculated to obtain the interest fluctuation value.

[0012] In a preferred embodiment, identifying whether a user has the indication that the electric vehicle used by the user participates in the V2G potential calculation refers to:

[0013] The average interest value is compared with the standard threshold one, and the interest fluctuation value is compared with the standard threshold two. If the average interest value is greater than or equal to the standard threshold one and the interest fluctuation value is less than or equal to the standard threshold two, a first-class signal is generated. If the average interest value is greater than or equal to the standard threshold one and the interest fluctuation value is less than or equal to the standard threshold two, a second-class signal is generated. The first-class signal indicates that the user has signs that the electric vehicle he uses participates in the V2G potential calculation, and the second-class signal indicates that the user does not have signs that the electric vehicle he uses participates in the V2G potential calculation.

[0014] In a preferred embodiment, the feature extraction content includes the time window potential feature combination in the user behavior, and the power potential feature combination in the user behavior. During the feature analysis, the time window potential index is generated based on the time window potential feature combination in the user behavior, and the power potential index is generated based on the power potential feature combination in the user behavior.

[0015] In a preferred embodiment, the logic for obtaining the time window potential index is:

[0016] The time window potential index calculation formula is:

[0017] θ is a preset non-zero adjustment factor, q is a preset nonlinear adjustment coefficient 1, which is less than 1, p is a preset nonlinear adjustment coefficient 2, which is greater than 1, C is a constant, TWP represents the time window potential index, Fstability represents the parking behavior regularity factor, Wpriority represents the grid demand matching priority factor, Ttotal represents the total parking time, and Toverlap represents the total overlapping time of the parking time and the grid demand time;

[0018] The calculation formula of parking behavior regularity factor Fstability is:

[0019] f(GL) represents the function of regularity GL, μduration represents the mean parking duration in historical parking data, σsatrt represents the standard deviation of parking start time in historical parking data, and σend represents the standard deviation of parking end time in historical parking data;

[0020] The calculation formula of the grid demand matching priority factor Wpriority is:

[0021] ti represents the overlapping time between the parking period and the i-th grid demand period, wi represents the preset priority coefficient of the i-th grid demand period, and the value range is [0.5, 1]. The higher the value, the more urgent the grid demand. n represents the number of grid demand periods.

[0022] In a preferred embodiment, the logic for obtaining the power potential index is:

[0023] The calculation formula of power potential index is:

[0024] f(TJ) represents the function of the nonlinear adjustment factor TJ, r1 and r2 are both preset non-zero proportional coefficients, and the sum of the two is less than or equal to one, Eusable represents the remaining power that can be used for V2G when the user parks, ΔEtrend represents the change value of the power during the user's parking period, Ereserve represents the power that the user expects to retain, Etotal represents the total capacity of the user's vehicle battery, and EPI represents the power potential index;

[0025] Eusable = max(0, Ecurrent - Efuture); Ecurrent is the current remaining power when parking, and Efuture is the power required to ensure travel;

[0026] ΔEtrend=Echarge-Edischarge; Echarge is the additional charge during parking, and Edischarge is the natural power loss during parking;

[0027] ΔR is the remaining power utilization rate deviation value, which is calculated by the following formula: Eneeded is the amount of electricity required by the grid during V2G operation.

[0028] In a preferred embodiment, classifying the user's travel habits according to the result of the feature analysis operation refers to:

[0029] The power potential index and the time window potential index are input into the pre-trained convolutional neural network model together. The convolutional neural network model outputs the classification results of the user's travel habits and divides the users into high-potential travel types, medium-potential travel types, and low-potential travel types.

[0030] In a preferred embodiment, the logic of fuzzy reasoning is:

[0031] The power potential index and time window potential index of users with medium potential travel type are taken as input variables, and the user's improvement direction is taken as the output variable. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the adaptability of the improvement direction under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to obtain the user's improvement direction.

[0032] Technical effects and advantages of the present invention:

[0033] The present invention quantifies the user's interest level in economic returns, green energy contribution and incentive mechanism through power supply willingness survey, and screens out user groups with potential willingness to participate in the V2G model. According to the average and fluctuation values ​​of user interest, the user's participation signs are accurately identified to ensure that the screened users are not only highly interested, but also have stable interests, which significantly improves the interaction effect between users and the power grid.

[0034] The Electric Power Potential Index (EPI) and Time Window Potential Index (TWP) are used to comprehensively evaluate the V2G participation capabilities of users in terms of power and time dimensions, providing a comprehensive potential analysis. The convolutional neural network (CNN) is used to classify users' travel habits and scientifically divide users into high-potential, medium-potential, and low-potential travel types, achieving accurate hierarchical management of user resources. Accurate potential assessment avoids resource waste and concentrates limited V2G resources on high-potential users.

[0035] For medium-potential users, fuzzy reasoning methods are used to derive specific optimization directions, and personalized improvement suggestions (optimizing power characteristics, optimizing time characteristics or comprehensive optimization) are provided to further enhance user potential. The present invention fully considers the actual travel needs and behavioral patterns of users, eliminates low-potential users, and ensures that the impact of V2G on users' normal travel is minimized. By tapping into the power potential and time potential of users, distributed electric vehicle energy storage resources are effectively integrated to provide the power grid with more flexible load regulation capabilities. Achieve power grid load balancing, peak and valley power optimization, and increase in renewable energy utilization, reduce the use of traditional fossil energy, and promote the development of green energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0037] Figure 1 This is a schematic diagram of a method for calculating the potential of electric vehicles to participate in V2G based on user travel habits in the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] Reference Figure 1 The following embodiments are obtained:

[0040] Example 1: The concept and significance of electric vehicles participating in V2G: The V2G mode refers to a two-way energy interaction mechanism between electric vehicles and the power grid. Electric vehicles not only obtain electrical energy from the power grid for their own charging, but can also return the remaining battery power to the power grid when parking to adjust the power load or cope with peak power demand. The potential for electric vehicles to participate in V2G depends on whether the user is willing and able to provide power support during non-driving periods. This mechanism helps to achieve renewable energy consumption and load balancing in the power grid, as well as improve the economic benefits of users. Based on this, the present invention proposes a method for calculating the potential of electric vehicles to participate in V2G based on user travel habits, including the following steps:

[0041] Conduct a survey on the user's willingness to supply power, then conduct a survey and analysis based on the survey results, and identify whether the user has signs of the electric vehicle he uses participating in the V2G potential calculation; by investigating the user's interest in economic returns, green energy, incentive mechanisms, etc., determine whether the user is willing to let his electric vehicle participate in the V2G mode. This is the starting point of the entire potential calculation, screening out users who are interested in V2G to avoid subsequent waste of resources and calculations. This step reflects the user's subjective willingness and is a prerequisite for participating in V2G. Use survey analysis technology to quantify the user's interest in V2G, and evaluate the stability and enthusiasm of the user's interest by calculating the average interest value and interest fluctuation value. By comparing with the preset threshold, a type I signal (with potential) or a type II signal (without potential) is generated to preliminarily screen users with the potential to participate in V2G, ensuring that the screened users are not only interested, but also have stable interest and willingness. Through the survey and analysis results, potential users are accurately identified to ensure that users entering the next step are worth further analysis. Through signal classification, clear user groups are formed. Type I signal users enter the subsequent feature analysis, and type II signal users are excluded to improve calculation efficiency.

[0042] If the user has signs that the electric vehicle he uses participates in the V2G potential calculation, the user's travel habits are sequentially obtained, feature extracted, and analyzed. Then, according to the results of the feature analysis operation, the user's travel habits are classified into high-potential travel types, medium-potential travel types, and low-potential travel types; basic data such as user parking behavior, charging behavior, and travel needs are obtained to form a user behavior data portrait to ensure that subsequent feature extraction and analysis have accurate data support and reflect real user habits. The accuracy of the data determines the reliability of potential calculation. The key elements in user behavior are extracted into a combination of time window potential features and a combination of power potential features. The time window feature reflects the user's parking time, regularity, and matching degree with the grid demand period, etc. The power feature reflects the user's remaining power utilization efficiency, power change trend, and willingness to retain power, etc. Through feature extraction, quantitative indicators are formed to facilitate subsequent analysis. Based on the results of feature extraction, the time window potential index and the power potential index are calculated to evaluate the time matching degree between the user's parking behavior and the grid demand and the power utilization potential, respectively, to form a clear numerical result for subsequent decision-making. The time window potential index and the power potential index are input into the convolutional neural network (CNN) model for classification, and the users are divided into high-potential travel type, medium-potential travel type and low-potential travel type. The purpose of classification is to optimize resource allocation and eliminate low-potential users who are not suitable for participating in V2G. High-potential users can participate directly, medium-potential users can be further optimized, and low-potential users will no longer enter the subsequent process.

[0043] Eliminate users of low-potential travel types, and use fuzzy reasoning to infer the improvement direction of users of medium-potential travel types. Eliminate low-potential users to ensure that resources are concentrated on high-potential and medium-potential users, avoid inefficient V2G operations or additional costs caused by low-potential users, and improve the overall V2G implementation efficiency. The fuzzy reasoning system optimizes and analyzes medium-potential users, infers their improvement directions (such as optimizing parking time or charging behavior), and generates an optimization strategy by inputting the user's time window potential index and power potential index and combining fuzzy rules. The goal of this step is to convert medium-potential users into high-potential users and further improve V2G participation and benefits.

[0044] The power supply willingness survey includes economic return interest survey, green energy interest survey, and incentive mechanism participation interest survey. Specific examples are:

[0045] Economic Return Interest Survey: Assess users’ interest in obtaining economic benefits through participating in V2G, such as electricity subsidies, revenue sharing, dynamic price adjustments and other incentives.

[0046] Survey question design: Are you willing to adjust your charging time to obtain electricity subsidies? If you participate in the V2G model, what is your minimum expectation of potential benefits? Will your willingness to participate be reduced due to limited benefits?

[0047] Purpose: To test the user's sensitivity to economic incentives, predict their actual acceptance of the V2G model, quantify the user's interest in revenue returns, and provide a basis for subsequent interest value calculations.

[0048] Green Energy Interest Survey: Assess users' interest in supporting green energy and environmental goals, such as reducing carbon emissions and promoting the use of renewable energy.

[0049] Survey question design: Are you willing to support the widespread use of green energy by participating in V2G? Are you willing to support environmental protection goals without direct economic benefits? Does participating in V2G conform to your environmental values?

[0050] Purpose: To test users’ potential willingness to participate under non-economic motivations and to quantify users’ interest in contributing to green energy, which is particularly suitable for user groups that are concerned about sustainable development.

[0051] Incentive mechanism participation interest survey: evaluate users' overall interest and acceptance of V2G incentive mechanisms (comprehensive economic, environmental, convenience and other factors).

[0052] Survey question design: Are you willing to participate in the V2G incentive program? (Overall willingness); If the incentive mechanism requires adjusting the charging time, are you able to accept it? Do you think participating in V2G will bring convenience or inconvenience to your life?

[0053] Purpose: To comprehensively evaluate users' overall interest and acceptance in participating in V2G incentive mechanisms and test whether users are willing to adjust their vehicle use and charging behaviors for multi-dimensional incentives.

[0054] According to the user's answer to each question, a score (such as 1-5 points) is assigned, and the average score of each type of question is summarized to calculate the corresponding score values ​​of the three types of interests: economic return interest value (E), green energy interest value (G), and incentive mechanism participation interest value (M). Through the investigation and analysis of the three types of interests, we can fully understand the sources of users' interest in V2G (economy, environmental protection, and comprehensive incentives), quantify users' interests, and form comparable standards. This will provide basic data for subsequent judgment of whether users have signs of V2G potential calculations, ensure the scientificity and accuracy of the user screening process, and improve the overall effect of the V2G plan.

[0055] Survey analysis refers to: obtaining the corresponding interest scores of users in the economic return interest survey, green energy interest survey, and incentive mechanism participation interest survey, and then calculating the average of the interest scores corresponding to the three types of interests to obtain the average interest value, and calculating the standard deviation of the interest scores corresponding to the three types of interests to obtain the interest fluctuation value. Identifying whether the user has signs that the electric vehicle he uses participates in the V2G potential calculation refers to: comparing the average interest value with the standard threshold one, and comparing the interest fluctuation value with the standard threshold two. If the average interest value is greater than or equal to the standard threshold one and the interest fluctuation value is less than or equal to the standard threshold two, a first type of signal is generated. If the average interest value is greater than or equal to the standard threshold one and the interest fluctuation value is less than or equal to the standard threshold two, a second type of signal is generated. The first type of signal indicates that the user has signs that the electric vehicle he uses participates in the V2G potential calculation, and the second type of signal indicates that the user does not have signs that the electric vehicle he uses participates in the V2G potential calculation.

[0056] Obtaining and quantifying interest scores: Calculate the user's scores in the economic return interest survey, green energy interest survey, and incentive mechanism participation interest survey, and convert qualitative questions into quantitative data. Provide data support for subsequent comprehensive analysis of interest values ​​to ensure that the multi-dimensional evaluation of user interests is more scientific and accurate.

[0057] Calculation of average interest value (overall interest level assessment): By calculating the average of the three types of interest score values, the user's overall interest in V2G is measured. The user's overall interest level is the basis for judging their potential to participate in V2G. Users above the threshold are more likely to actually participate.

[0058] Calculation of interest fluctuation value (stability assessment): By calculating the standard deviation of the three types of interest scores, the stability of user interests is assessed (whether the preferences are concentrated, whether there are contradictions or uncertainties). Users with higher stability are more worthy of further investment in potential analysis, while users with large interest fluctuations may not be suitable for the V2G model.

[0059] Quantify interest level and stability: The average and fluctuation values ​​of interest values ​​provide a scientific quantitative basis for subsequent signal classification, avoiding subjective judgment. The user interest levels are clearly divided into categories that are potentially suitable for participating in V2G and those that are not suitable for participating in V2G, improving screening efficiency and accuracy.

[0060] The comparison between the average interest value and the standard threshold 1 determines whether the user's overall interest meets the minimum requirement, ensuring that the selected users have a clear willingness to participate. The comparison between the interest fluctuation value and the standard threshold 2 filters out users with large interest fluctuations. Such users have unstable interests and low long-term benefits from participating in V2G. Through the comparison of the two dimensions, qualified users are accurately screened out to improve the pertinence of computing and resource investment. The first type of signal indicates that the user has signs of V2G potential and enters the next step of travel habit analysis. The second type of signal indicates that the user does not have signs of V2G potential and does not enter the subsequent steps to save computing resources. Signal classification, as a hierarchical screening mechanism, excludes unsuitable users, avoids subsequent over-analysis of low-willing users, and optimizes the overall process efficiency. Users who do not have signs do not need to enter the subsequent feature analysis and classification links, control the computing burden from the source, accurately screen potential user groups, maximize the utilization efficiency of V2G resources, and screen out users who are not only highly interested but also have stable behaviors, which helps to form a stable V2G user base. The dual comparison of interest value and fluctuation value builds an efficient user selection mechanism, lays the foundation for the success of the V2G participation plan, and reduces the number of dropouts.

[0061] The content of feature extraction includes the time window potential feature combination in user behavior and the power potential feature combination in user behavior. During feature analysis, a time window potential index is generated based on the time window potential feature combination in user behavior, and a power potential index is generated based on the power potential feature combination in user behavior.

[0062] The time window potential index (TWP) is used to measure the potential of user parking behavior to participate in V2G in the time dimension, reflecting the matching degree between the parking period and the grid demand period, the regularity of parking behavior and the contribution of total parking time. The larger the index value, the more suitable the user's parking behavior is for participating in V2G and the higher the potential. The logic for obtaining the time window potential index is: The calculation formula for the time window potential index is: θ is a preset non-zero adjustment factor, which controls the distribution range of the final time window potential index, making the time window potential index more sensitive to changes in important parameters. q is a preset nonlinear adjustment coefficient of one, which is less than 1, weakening the influence of the total duration on the time window potential index. p is a preset nonlinear adjustment coefficient of two, which is greater than 1, enhancing the contribution of Toverlap to the time window potential index. C is a constant. TWP represents the time window potential index. Fstability represents the regularity factor of parking behavior. Wpriority represents the grid demand matching priority factor. Ttotal represents the total parking duration, which is used for normalization processing to avoid excessive influence of the total parking duration on the index. The longer the total duration, the more opportunities to participate in V2G may be provided. Toverlap represents the total overlap time between parking time and grid demand time, reflecting the matching degree between user parking behavior and grid demand. The longer the overlap time, the more the user parking behavior meets the grid demand.

[0063] The parking behavior regularity factor Fstability measures the time regularity of the user's parking behavior. The stronger the regularity, the higher the V2G potential. The calculation formula of the parking behavior regularity factor Fstability is:

[0064] f(GL) represents the function of regularity GL, μduration represents the mean parking duration in historical parking data, σsatrt represents the standard deviation of parking start time in historical parking data, and σend represents the standard deviation of parking end time in historical parking data;

[0065] The grid demand matching priority factor Wpriority reflects the matching degree between parking behavior and the high-priority demand period of the grid. The more urgent the demand, the larger the priority factor value. The calculation formula of the grid demand matching priority factor Wpriority is:

[0066] ti represents the overlapping time between the parking period and the i-th grid demand period, wi represents the preset priority coefficient of the i-th grid demand period, and its value range is [0.5, 1]. The higher the value, the more urgent the grid demand. n represents the number of grid demand periods, reflecting the complexity of the overall grid demand.

[0067] The Electricity Potential Index (EPI) is used to evaluate the potential of users to participate in V2G at the electricity level. By comprehensively considering the remaining electricity, electricity change trend, willingness to retain electricity, and the matching degree of electricity utilization, the formula reflects the adaptability of the user's electricity characteristics to V2G. The design of the formula uses multi-dimensional features and nonlinear functions to make the results more accurate and dynamic. The logic for obtaining the Electricity Potential Index is:

[0068] The calculation formula of power potential index is:

[0069] f(TJ) represents the functional formula of the nonlinear adjustment factor TJ, r1 and r2 are both preset non-zero proportional coefficients, and the sum of the two is less than or equal to one, which regulates the influence ratio of static power and dynamic power to ensure the flexibility of the formula in different scenarios. Eusable represents the remaining power that can be used for V2G when the user is parking, ΔEtrend represents the change value of the power during the user's parking period, Ereserve represents the power that the user expects to retain, Etotal represents the total capacity of the user's vehicle battery, and EPI represents the power potential index, which reflects the possibility of using the user's remaining power and its matching degree with the power grid demand. The numerator represents the effective power currently available in the user's electric vehicle. The static power and dynamic adjustments (such as trends and power reserves) are comprehensively considered to construct a comprehensive evaluation of the user's power, focusing on both the actual value and the behavioral preferences and dynamic changes. The denominator is normalized to eliminate the comparison deviation caused by the different total battery capacities, provide a standardized perspective on the power ratio, and make the results comparable across users. The nonlinear adjustment function corrects the power potential index, considers the deviation between the remaining power and the demand, and enables the formula to dynamically adjust the results under different circumstances, ensuring the flexibility of the potential index and making it closer to the actual needs of the power grid;

[0070] Eusable = max(0, Ecurrent - Efuture); Ecurrent is the current remaining power when parking, and Efuture is the power required to ensure travel;

[0071] ΔEtrend=Echarge-Edischarge; Echarge is the additional charge during parking, and Edischarge is the natural power loss during parking; ΔEtrend is the power change during parking, taking into account the additional charge and power loss. If ΔEtrend is greater than zero, it means that the user has additional power charged during parking, increasing the V2G potential. If ΔEtrend is less than zero, it means that the power loss is large and the potential is reduced.

[0072] The S-curve (Logistic function) is used to adjust the power potential. The potential index is dynamically adjusted through the function to avoid unreasonable evaluation caused by excessive power deviation. λ is the preset sensitivity adjustment coefficient, which is greater than zero. ΔR is the remaining power utilization rate deviation value, which is calculated by the following formula: Eneeded is the amount of electricity required by the grid during V2G operation. ΔR>0 means that the user’s remaining electricity is more than the grid demand, and the potential is higher; ΔR<0 means that the user’s remaining electricity is insufficient to meet the grid demand, and the potential is lower.

[0073] Classifying the user's travel habits based on the results of the feature analysis operation means:

[0074] The power potential index and the time window potential index are input into the pre-trained convolutional neural network model together. The convolutional neural network model outputs the classification results of the user's travel habits and divides the users into high-potential travel types, medium-potential travel types, and low-potential travel types.

[0075] Input features: Electricity potential index (EPI): It indicates the electricity potential that users can use for V2G during parking. It combines features such as available electricity, dynamic trends, and grid demand matching. The higher the value, the stronger the ability to support V2G at the electricity level.

[0076] Time Window Potential Index (TWP): Indicates the adaptability of the time characteristics of user parking behavior to V2G, reflecting the degree of overlap between the parking period and the grid demand period, behavioral regularity, etc. The higher the value, the more suitable the time level is for participating in V2G.

[0077] Pre-trained CNN model: Model structure: The convolutional neural network extracts local patterns and global correlations of input features through convolutional layers, and uses fully connected layers to classify users.

[0078] Specifically include: Convolution layer: extract the potential correlation pattern between EPI and TWP (such as the joint characteristics of power and time matching). Pooling layer: downsample to extract important features and reduce the amount of calculation. Fully connected layer: combine multiple layers of features to generate classification results. Output layer: output probability value (the possibility of the user belonging to a certain type).

[0079] Training data: Use a large amount of historical data of users (EPI, TWP, and corresponding actual participation) to train the model. Training goal: The model can accurately predict the user's travel habit type based on the input features. Model output: CNN outputs the probability value of the user belonging to each category, and the category corresponding to the maximum probability value is the final classification result.

[0080] The logic of fuzzy reasoning is: take the power potential index and time window potential index of users with medium potential travel type as input variables, take the user's improvement direction as output variable, fuzzify the input variables, convert the input variable value into fuzzy set, fuzzify the output variables, convert the output variables into fuzzy set, formulate fuzzy rules, describe the adaptability of improvement direction under different data type combinations, reason with the fuzzified input variables through fuzzy rules, and obtain the user's improvement direction.

[0081] Input variables include: Electricity potential index (EPI): Indicates the user's participation potential at the electricity level. The value range is usually [0,1]. Low values ​​indicate insufficient electricity potential (such as insufficient remaining electricity or poor electricity trend), and high values ​​indicate good electricity potential. Time window potential index (TWP): Indicates the matching degree between the time dimension of the user's parking behavior and the grid demand. The value range is usually [0,1]. Low values ​​indicate that the parking time has little overlap with the grid demand or the parking regularity is poor, and high values ​​indicate high time potential.

[0082] The output variable is the improvement direction of the user, and the results include the following three types:

[0083] Optimize power characteristics: Improve the user's power potential, such as increasing the remaining power when parking, reducing the reserved power or additional charging behavior during parking.

[0084] Optimize time characteristics: Improve the user's time window potential, such as adjusting the parking time period or improving the matching of parking periods with grid demand.

[0085] Comprehensive optimization: Improves both power and time characteristics at the same time, suitable for users whose potential for both is not high but still has room for improvement.

[0086] Convert the values ​​of EPI and TWP into fuzzy sets and define fuzzy membership functions. For example: Fuzzy set of EPI: low, medium, high. Fuzzy set of TWP: low, medium, high. Fuzzy sets of output variables include: optimizing power characteristics, optimizing time characteristics, and comprehensive optimization. Formulation of fuzzy rules: Formulate a set of fuzzy rules based on the user's EPI and TWP value combination. The examples are as follows: Rule 1: If EPI is low and TWP is low, then optimize comprehensively. Rule 2: If EPI is low and TWP is high, then optimize power characteristics. Rule 3: If EPI is high and TWP is low, then optimize time characteristics. Rule 4: If EPI is medium and TWP is medium, then optimize comprehensively. The rules describe the best improvement direction under different potential index combinations.

[0087] Use reasoning methods to match the fuzzy input values ​​with the rule base, derive the output fuzzy set, and convert the inferred fuzzy set into a specific output value through defuzzification methods (such as the centroid method and the maximum membership method) to determine the user's improvement direction.

[0088] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0089] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0090] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0091] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0092] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for calculating the potential of electric vehicles to participate in V2G based on user travel habits, characterized in that: The following steps are involved: Conduct a survey on users' willingness to supply electricity, then conduct a survey and analysis based on the survey results, and identify whether the users have signs that their electric vehicles can participate in the V2G potential calculation; If the user has signs that the electric vehicle used by the user participates in the V2G potential calculation, the user's travel habits are sequentially subjected to information acquisition, feature extraction, and feature analysis operations, and then the user's travel habits are classified according to the results of the feature analysis operations, and the user is divided into a high potential travel type, a medium potential travel type, and a low potential travel type; Users with low potential travel types are eliminated, and fuzzy reasoning is used to infer the improvement direction of users with medium potential travel types.

2. The method for calculating the potential of electric vehicles to participate in V2G based on user travel habits according to claim 1, characterized in that: The power supply willingness survey includes an economic return interest survey, a green energy interest survey, and an incentive mechanism participation interest survey.

3. The method for calculating the potential of electric vehicles to participate in V2G based on user travel habits according to claim 2, characterized in that: Research analysis refers to: The corresponding interest scores of users in the economic return interest survey, green energy interest survey, and incentive mechanism participation interest survey are obtained respectively, and then the average of the interest score values ​​corresponding to the three types of interests is calculated to obtain the average interest value, and the standard deviation of the interest score values ​​corresponding to the three types of interests is calculated to obtain the interest fluctuation value.

4. The method for calculating the potential of electric vehicles to participate in V2G based on user travel habits according to claim 3, characterized in that: Indications for identifying whether a user has the potential to participate in V2G calculations using an electric vehicle include: The average interest value is compared with the standard threshold one, and the interest fluctuation value is compared with the standard threshold two. If the average interest value is greater than or equal to the standard threshold one and the interest fluctuation value is less than or equal to the standard threshold two, a first-class signal is generated. If the average interest value is greater than or equal to the standard threshold one and the interest fluctuation value is less than or equal to the standard threshold two, a second-class signal is generated. The first-class signal indicates that the user has signs that the electric vehicle he uses participates in the V2G potential calculation, and the second-class signal indicates that the user does not have signs that the electric vehicle he uses participates in the V2G potential calculation.

5. The method for calculating the potential of electric vehicles to participate in V2G based on user travel habits according to claim 4, characterized in that: The content of feature extraction includes the time window potential feature combination in user behavior and the power potential feature combination in user behavior. During feature analysis, a time window potential index is generated based on the time window potential feature combination in user behavior, and a power potential index is generated based on the power potential feature combination in user behavior.

6. The method for calculating the potential of electric vehicles to participate in V2G based on user travel habits according to claim 5, characterized in that: The logic for obtaining the time window potential index is: The time window potential index calculation formula is: θ is a preset non-zero adjustment factor, q is a preset nonlinear adjustment coefficient 1, which is less than 1, p is a preset nonlinear adjustment coefficient 2, which is greater than 1, C is a constant, TWP represents the time window potential index, Fstability represents the parking behavior regularity factor, Wpriority represents the grid demand matching priority factor, Ttotal represents the total parking time, and Toverlap represents the total overlapping time of the parking time and the grid demand time; The calculation formula of parking behavior regularity factor Fstability is: f(GL) represents the function of regularity GL, μduration represents the mean parking duration in historical parking data, σsatrt represents the standard deviation of parking start time in historical parking data, and σend represents the standard deviation of parking end time in historical parking data; The calculation formula of the grid demand matching priority factor Wpriority is: ti represents the overlapping time between the parking period and the i-th grid demand period, wi represents the preset priority coefficient of the i-th grid demand period, and the value range is [0.5, 1]. The higher the value, the more urgent the grid demand. n represents the number of grid demand periods.

7. The method for calculating the potential of electric vehicles to participate in V2G based on user travel habits according to claim 6, characterized in that: The logic for obtaining the power potential index is: The calculation formula of power potential index is: f(TJ) represents the function of the nonlinear adjustment factor TJ, r1 and r2 are both preset non-zero proportional coefficients, and the sum of the two is less than or equal to one, Eusable represents the remaining power that can be used for V2G when the user parks, ΔEtrend represents the change value of the power during the user's parking period, Ereserve represents the power that the user expects to retain, Etotal represents the total capacity of the user's vehicle battery, and EPI represents the power potential index; Eusable = max(0, Ecurrent - Efuture); Ecurrent is the current remaining power when parking, and Efuture is the power required to ensure travel; ΔEtrend=Echarge-Edischarge; Echarge is the additional charge during parking, and Edischarge is the natural power loss during parking; λ is the preset sensitivity adjustment coefficient, which is greater than zero. ΔR is the remaining power utilization rate deviation value, which is calculated by the following formula: Eneeded is the amount of electricity required by the grid during V2G operation.

8. The method for calculating the potential of electric vehicles to participate in V2G based on user travel habits according to claim 7, characterized in that: Classifying the user's travel habits based on the results of the feature analysis operation means: The power potential index and the time window potential index are input into the pre-trained convolutional neural network model together. The convolutional neural network model outputs the classification results of the user's travel habits and divides the users into high-potential travel types, medium-potential travel types, and low-potential travel types.

9. The method for calculating the potential of electric vehicles to participate in V2G based on user travel habits according to claim 8, characterized in that: The logic of fuzzy reasoning is: The power potential index and time window potential index of users with medium potential travel type are taken as input variables, and the user's improvement direction is taken as the output variable. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the adaptability of the improvement direction under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to obtain the user's improvement direction.

Images