Electric vehicle electric quantity state and remaining mileage prediction method

Abstract

The invention provides an electric vehicle electric quantity state and remaining mileage prediction method, which comprises the following steps: carrying out regional division on a battery pack, deploying a sensor, collecting electric quantity data of each region in real time, constructing an anomaly detection model to identify an abnormal region, carrying out local data adjustment, and combining user driving behaviors and environmental parameters to predict the electric quantity state and remaining mileage of an electric vehicle. A long-short-term memory network is adopted to dynamically predict the power consumption trend, the remaining mileage is calculated according to real-time road conditions and temperature changes, meanwhile, an adaptive correction mechanism is introduced, and model parameters are dynamically adjusted through comparative analysis with historical prediction records, so that the prediction accuracy is improved. According to the method, accurate monitoring of the electric quantity state of the electric vehicle and dynamic prediction of the remaining mileage can be achieved, the problem that a traditional method is large in prediction deviation in a complex driving environment is effectively solved, more reliable electric quantity and mileage information is provided for a user, and the use experience and safety of the electric vehicle are improved.

Description

Technical Field

[0001] The invention relates to a method for predicting the power state and remaining mileage of an electric vehicle. Background Art

[0002] As an important pillar of modern transportation and energy transformation, the new energy vehicle sector has irreplaceable value in promoting green travel and reducing carbon emissions. Its core technology directly affects vehicle performance and user experience. However, current solutions for power monitoring and prediction still have obvious deficiencies. In actual use, many methods often have low accuracy, resulting in large deviations between the power display and the actual situation, which in turn affects the user's judgment of the remaining mileage and reduces the reliability and convenience of driving.

[0003] In this context, the accuracy of power monitoring has become a key issue that needs to be addressed. The primary challenge is that when the battery pack is taken as a whole, the collection of power data often lacks detailed management, making it difficult to capture abnormal changes in local areas, resulting in errors in the overall power estimation. This limitation of holistic monitoring further leads to another problem, namely, the inability to make personalized adjustments based on the usage habits and driving environment of different users, and the difference between theoretical power consumption and actual consumption cannot be effectively corrected, which often makes the remaining mileage prediction inaccurate. These two problems are interrelated. The former leads to inaccurate data basis, while the latter exacerbates the impact of errors due to the lack of a personalized correction mechanism. Summary of the invention

[0004] The present invention proposes a method for predicting the power state and remaining mileage of an electric vehicle, which provides users with more reliable power and mileage information by refining battery pack management and building a personalized power correction mechanism in combination with users' actual usage scenarios.

[0005] The technical solution of the present invention is achieved in this way:

[0006] A method for predicting the state of charge and remaining mileage of an electric vehicle, the method comprising:

[0007] S1, by dividing the battery pack into regions, building a multi-point power data collection framework, collecting the power status data of each region in real time in each region, and forming a regional power original data set; analyzing the voltage and current data of each region one by one, if the data of a certain region exceeds the preset threshold range, it is marked as an abnormal region, and the abnormal region distribution information is generated; according to the abnormal region distribution information, a local power correction matrix is constructed, and the power data of the abnormal region is weighted and adjusted, and combined with the data of the non-abnormal region, the power estimation value of the entire battery pack is recalculated to obtain the corrected overall power status data; S2. Combine the user's historical driving records and driving environment parameters obtained in real time, including the acceleration frequency, average vehicle speed, and road slope, to construct a user behavior feature vector; according to the user behavior feature vector and the corrected overall battery state data, use a long short-term memory network model to dynamically predict the power consumption trend and generate power consumption rate prediction data under specific scenarios; S3. Combine the current driving environment adaptation parameters, including real-time road conditions and temperature changes, and calculate the remaining mileage prediction data through a preset mileage conversion formula: remaining mileage = (corrected overall battery state data / power consumption rate prediction data) * environment adjustment coefficient, where the environment adjustment coefficient is obtained from a preset table based on temperature and road conditions, and finally obtain the preliminary remaining mileage prediction data; S4. For the preliminary remaining mileage prediction data, conduct a comparative analysis with the recent historical prediction records stored in the system. If the prediction deviation exceeds the preset threshold range, trigger an adaptive correction mechanism to adjust the parameters of the long short-term memory network model and generate updated model parameters; combine the latest regional battery raw data set and the user behavior feature vector to re-predict the power consumption rate and generate optimized power consumption rate prediction data; S5. For the optimized power consumption rate prediction data, continuously update the corrected overall battery state data and driving environment adaptation parameters, and dynamically adjust the remaining mileage prediction data through iterative calculations with a set maximum number of iterations or convergence conditions, and finally determine the battery state and mileage prediction data for real-time display.

[0008] The beneficial effects of the present invention are as follows: By dividing the battery pack into regions and deploying sensors, collecting the power data of each region in real time, constructing an anomaly detection model to identify the abnormal regions and perform local data adjustment, combining the user's driving behavior and environmental parameters, using a long short-term memory network to dynamically predict the power consumption trend, calculating the remaining mileage according to real-time road conditions and temperature changes, and introducing an adaptive correction mechanism to dynamically adjust the model parameters through comparative analysis with historical prediction records, the prediction accuracy is improved. The present invention can achieve accurate monitoring of the battery state of electric vehicles and dynamic prediction of the remaining mileage, effectively solve the problem of large prediction deviation of traditional methods in complex driving environments, provide more reliable battery and mileage information for users, and improve the use experience and safety of electric vehicles. Detailed implementation manners

[0009] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0010] A method for predicting the state of charge and remaining mileage of an electric vehicle, specifically including:

[0011] Step S101: By dividing the battery pack into regions, a multi-point power data acquisition framework is constructed. Independent voltage sensors and current sensors are deployed in each region to collect the power state data of each region in real time, forming a raw data set of regional power, laying a data foundation for subsequent anomaly detection and overall power estimation.

[0012] According to the business attributes of the regional division, the battery pack is managed by region. Voltage sensors and current sensors are deployed in each region to monitor the power state of each region in real time. Voltage values and current values are obtained from the sensors to obtain an initial data stream divided by region. Through the business attributes of data acquisition, the initial data stream divided by region is transmitted to a pre-established data storage unit, and the voltage values and current values from different regions are integrated and processed by means of data fusion to determine the integrity of the power state of each region, and a unified comprehensive data set is obtained. If the power state data of a certain region in the comprehensive data set exceeds the preset threshold range, then for the business attributes of anomaly detection, a preset logical judgment module is called to compare and analyze the voltage value and current value of the region to determine whether there is a potential anomaly. According to the business attributes of the overall power, the power state data of all regions are extracted from the comprehensive data set, and combined with the business attributes of state evaluation, the data are comprehensively processed by means of weighted average calculation to obtain an estimated value of the overall power of the battery pack.

[0013] Step S102: For the raw data set of regional power, apply a pre-established anomaly detection model to analyze the voltage and current data of each region one by one. If the data of a certain region exceeds the preset threshold range, it is marked as an abnormal region, and abnormal region distribution information is generated for subsequent local data adjustment.

[0014] According to the pre-established anomaly detection rules, voltage and current data of each region are obtained from the raw sub-region power consumption dataset, and each data point is compared one by one. If the voltage or current of a certain region exceeds the preset threshold range, it is marked as an abnormal region, and the preliminary abnormal region distribution information is obtained. By performing a secondary verification on the preliminary abnormal region distribution information, the adjacent region data of each abnormal region is obtained, and the mean comparison method is used to determine whether the voltage and current of the adjacent regions deviate. If the adjacent region data also exceeds the preset threshold range, it is determined that the abnormal region mark is valid, and the confirmed abnormal region list is obtained. According to the confirmed abnormal region list, the time series data of the abnormal regions are extracted from the raw dataset, the fluctuation amplitude is calculated for the time series data, and a standard deviation calculation tool is used to determine whether the fluctuation continuously exceeds the preset threshold range, obtaining the fluctuation characteristic distribution of the abnormal regions. By classifying the fluctuation characteristic distribution of the abnormal regions, the fluctuation characteristic categories are obtained, and a data smoothing tool is used for local adjustment for the categories, and it is determined whether the adjusted data points return to the preset threshold range, generating the final local data adjustment plan.

[0015] Step S103, according to the abnormal region distribution information, construct a local power consumption correction matrix, perform weighted adjustment on the power consumption data of the abnormal regions, and combine the data of the non-abnormal regions to recalculate the power consumption estimation value of the overall battery pack, obtaining the corrected overall power state data, providing an accurate basis for subsequent consumption prediction.

[0016] According to the abnormal regions and distribution information, construct a preliminary mapping table for identifying power consumption anomalies, obtain the power consumption data of each region from the battery configuration, and use a preset threshold for comparison for the division of the abnormal regions and non-abnormal regions, obtaining the determination results of the specific locations and influence ranges of the abnormal regions. Through the determination results of the abnormal regions, construct a correction matrix for local power consumption, perform weighted adjustment on the power consumption data of the abnormal regions, obtain the adjusted local power consumption values, and integrate them with the power consumption data of the non-abnormal regions to determine the preliminary overall power consumption distribution table. Using the overall power consumption distribution table, combine the operating parameters of the battery configuration to perform secondary calibration on the overall power consumption. If the calibrated power consumption estimation value does not match the preset threshold range, fine-tune the correction matrix to obtain the final overall power state data. According to the overall power state data, generate a basic dataset for consumption prediction, extract features for key indicators, obtain the input parameters required for prediction, determine whether the accuracy requirements are met, and determine the final power consumption estimation value.

[0017] Step S104: For the corrected overall power state data, combine the user's historical driving records and driving environment parameters obtained in real time, including acceleration frequency, average vehicle speed, and road gradient, to construct a user behavior feature vector for subsequent personalized power consumption trend analysis.

[0018] Obtain the acceleration frequency, average vehicle speed, and braking frequency data from the user's historical driving records. At the same time, extract the road gradient and environmental temperature data from the driving environment parameters. Denoise the data using a data cleaning tool to obtain an organized driving behavior data set. According to the driving behavior data set, use a data fusion tool to perform correlation mapping between the acceleration frequency and the average vehicle speed, and between the braking frequency and the road gradient to determine the driving habit characteristic values of the user in different environments. Through a pre-established calculation rule, obtain the driving habit characteristic values and the environmental temperature data. If the characteristic values exceed the preset threshold, dynamically adjust the power state data and determine the adjusted power consumption baseline value.

[0019]

[0020] DH represents the driving habit characteristic value, and T represents the total number of sampling time periods. represents the acceleration at time t. represents the braking force at time t. represents the steering angle at time t. This formula calculates the comprehensive characteristic value of driving behavior.

[0021]

[0022] ET represents the environmental temperature influence factor. represents the environmental temperature data, and α, β, and γ are preset temperature influence coefficients. This formula describes the influence relationship of environmental temperature on power consumption.

[0023]

[0024] δ represents the judgment result of the driving habit exceeding the threshold, and DH represents the driving habit characteristic value. represents the preset driving habit threshold. When the characteristic value exceeds the threshold, power adjustment is required.

[0025]

[0026] represents the adjusted power consumption baseline value. represents the basic power consumption value. represents the flag indicating whether adjustment is required. represents the driving habit adjustment coefficient, and DH represents the driving habit characteristic value. Let \(\alpha\) represent the temperature influence adjustment coefficient, and \(ET\) represent the environmental temperature influence factor. This formula realizes the dynamic adjustment of power consumption according to driving habits and environmental temperature.

[0027] For the power consumption reference value, a time series tool is used to conduct correlation analysis on the reference value and the driving mileage data to obtain the user's personalized power consumption trend characteristic value.

[0028] Step S105: According to the user behavior feature vector and the corrected overall power state data, use a long short-term memory network model to dynamically predict the power consumption trend, generate the power consumption rate prediction data under a specific scenario, and provide the core parameters for the subsequent remaining mileage calculation.

[0029] According to the user behavior feature vector and the corrected overall power state data, perform standardization processing through a pre-established data processing module to obtain the unified feature data set, and filter out outliers using a preset threshold during processing to obtain the cleaned basic data set. Using the cleaned basic data set, segment the power consumption historical records through a time series decomposition tool to obtain the consumption pattern characteristics within different time periods. If the consumption fluctuation in a certain time period exceeds the preset threshold, smooth the data segment to determine the intermediate feature set. Through the intermediate feature set, dynamically model the power consumption trend, obtain the consumption rate prediction value under the specific scenario, and adjust the weights of different scenario features during processing to judge the prediction rate data that matches the user behavior. According to the prediction rate data, combine the user's current behavior environment information through a scenario adaptation tool to perform secondary calibration on the prediction value to obtain the final power consumption rate prediction result.

[0030] Step S106: For the power consumption rate prediction data, combine the current driving environment adaptation parameters, including real-time road conditions and temperature changes, and calculate the remaining mileage prediction data through a preset mileage conversion formula. The formula is: the remaining mileage is equal to the corrected overall power state data divided by the power consumption rate prediction data and then multiplied by the environmental adjustment coefficient. The environmental adjustment coefficient is obtained from a preset table based on temperature and road conditions. The remaining mileage represents the distance that the vehicle can continue to travel, the corrected overall power state data represents the current available power of the battery pack, the power consumption rate prediction data represents the speed of power consumption per unit time, and the environmental adjustment coefficient represents the influence factor of the environment on power consumption. Finally, obtain the preliminary remaining mileage prediction data.

[0031] According to the real-time road condition information and temperature change data, query the records matching the road conditions and the temperature from the preset environment table to determine the value of the environmental adjustment coefficient. By obtaining the corrected power status data and combining with the actual situation of the battery pack power, use the internal calibration tool to calibrate the current available power to obtain the accurate corrected power status value. If both the corrected power status value and the power consumption rate prediction data have been obtained, calculate through the mileage conversion formula, that is, the remaining mileage is equal to the corrected power status value divided by the power consumption rate prediction data and then multiplied by the environmental adjustment coefficient to obtain the preliminary remaining mileage prediction result. According to the preliminary remaining mileage prediction result, combined with the driving distance estimation logic, use the data comparison tool to perform a secondary verification on the result, judge whether it meets the reasonable range of consumption per unit time, and determine the final remaining mileage prediction data.

[0032] Step S107, for the preliminary remaining mileage prediction data, compare and analyze it with the recent historical prediction records stored in the system. If the prediction deviation exceeds the preset threshold range, trigger the adaptive correction mechanism to adjust the parameters of the long short-term memory network model and generate updated model parameters for improving the subsequent prediction accuracy.

[0033] Obtain the remaining mileage prediction data of the recent records from the repository, compare each item of the recent records with the currently generated prediction data, and calculate the difference value of each group of data to obtain the specific value of the prediction deviation. If the prediction deviation exceeds the preset threshold range, trigger the adaptive correction mechanism, and determine the direction and amplitude of the parameters to be adjusted by comparing the deviation trend in the historical records. According to the determined direction and amplitude of the parameters, use the pre-established adjustment rules to update the parameters of the long short-term memory network and generate the updated parameter configuration.

[0034]

[0035] Among them represents the currently generated predicted mileage data, represents the actual mileage data of the recent records obtained from the repository, represents the difference value between the two, that is, the specific value of the prediction deviation.

[0036]

[0037] Among them represents the preset threshold range, n represents the number of historical records, represents the prediction deviation value in the historical records, represents the standard deviation of the deviation, represents the adjustment coefficient;

[0038]

[0039] wherein represents the direction of parameter adjustment, m represents the number of records in the most recent period, represents the prediction deviation;

[0040]

[0041] wherein represents the final adjustment amplitude, represents the current prediction deviation, represents the average value of historical deviations, and represent the maximum and minimum values of historical deviations respectively, represents the adjustment coefficient;

[0042]

[0043] wherein represents the updated parameter, represents the original parameter, represents the direction of parameter adjustment, represents the adjustment amplitude, represents the gradient of the parameter. Through this formula, the system can update the parameters of the long short-term memory network according to the determined parameter direction and amplitude, generating an updated parameter configuration.

[0044] By comparing the average deviation of the most recent m records with the average deviation of the previous m records, the change trend of the deviation is determined, thereby deciding the direction of parameter adjustment. When the current prediction deviation exceeds this range, the adaptive correction mechanism will be triggered. By applying the updated parameter configuration to the prediction process, new remaining mileage prediction data is obtained, and compared with the historical records again to determine whether the prediction deviation has been reduced to within the threshold range.

[0045] Step S108, according to the updated model parameters, combined with the latest regional electricity consumption original data set and user behavior feature vectors, re-predict the electricity consumption rate, generating optimized electricity consumption rate prediction data to provide more accurate parameter support for subsequent remaining mileage prediction.

[0046] According to the latest regional power consumption raw dataset, obtain the historical records of power consumption in each region. Combine with the user behavior feature vectors, and use a data cleaning tool to denoise the original data to obtain the sorted basic power consumption dataset. For the sorted basic power consumption dataset, use the pre-established user behavior influence weight table to obtain the influence coefficient of each behavior feature on power consumption, and use a weighted calculation tool to determine the preliminary prediction value of the power consumption rate in each region.

[0047]

[0048] Among them represents the preliminary prediction value of the power consumption rate in region i, represents the influence weight coefficient of behavior feature j, represents the value of behavior feature j in region i, and n represents the total number of behavior features.

[0049] By weighted summing of each behavior feature, the preliminary prediction value is obtained. If the deviation between the preliminary prediction value of the power consumption rate and the historical true value exceeds the preset threshold, use a regression analysis tool to correct the preliminary prediction value to obtain the adjusted power consumption rate prediction dataset.

[0050]

[0051] Among them represents the prediction deviation percentage in region i, represents the preliminary prediction value of the power consumption rate in region i, represents the historical true consumption rate in region i. When exceeds the preset threshold, correction is required;

[0052]

[0053] Among them, CR_i' represents the predicted value of the power consumption rate after correction in region i, represents the preliminary prediction value, HR_i represents the historical true value, represents the correction coefficient, and its value range is from 0 to 1, which is used to control the degree of correction.

[0054] When exceeds the preset threshold, correction is required. According to the adjusted power consumption rate prediction dataset, combine with the geographical environment data of each region, and use a data fusion tool to perform secondary optimization on the prediction dataset to judge the final prediction result of the power consumption rate.

[0055]

[0056] Among them Indicates the predicted result of the final power consumption rate in area i, Indicates the corrected predicted value, Indicates the influence value of the k-th geographical environment factor in area i, Indicates the weight coefficient of the k-th geographical environment factor, and m represents the total number of geographical environment factors. The prediction result is optimized twice by integrating geographical environment data.

[0057] Step S109: For the optimized power consumption rate prediction data, continuously update the corrected overall power state data and driving environment adaptation parameters. Through loop iteration calculations with a set maximum number of iterations or convergence conditions, dynamically adjust the remaining mileage prediction data, and finally determine the power state and mileage prediction data for real-time display.

[0058] According to the power consumption data and the driving environment data, obtain the corresponding adaptation parameters from a pre-established repository. By comparing the current power consumption rate with historical records, determine whether the adaptation parameters are within a preset threshold range. If they exceed the threshold range, correct the adaptation parameters to obtain an adjusted parameter value. Using the adjusted parameter value, combined with the power state data, through loop iteration calculation methods, limit the maximum number of iterations and convergence conditions, and dynamically adjust the remaining mileage prediction data during each iteration to determine a preliminary mileage prediction result. By comparing the preliminary mileage prediction result with the power state data for real-time display, if the deviation exceeds a preset threshold, re-obtain the latest driving environment data and power consumption data, update the adaptation parameters, and obtain a corrected mileage prediction value. According to the corrected mileage prediction value and the power state data, combined with real-time display requirements, integrate the final display content through data fusion, and determine whether it meets the accuracy requirements. If not, return to the loop iteration calculation link to obtain more accurate display data.

[0059] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for predicting the state of charge and remaining mileage of an electric vehicle, characterized in that, The method includes: S1. By dividing the battery pack into regions, a multi-point power data acquisition framework is constructed. The power status data of each region is collected in real time in each region to form a raw data set of regional power. The voltage and current data of each region are analyzed one by one. If the data of a certain region exceeds the preset threshold range, it is marked as an abnormal region, and the abnormal region distribution information is generated. According to the abnormal region distribution information, a local power correction matrix is constructed to perform weighted adjustment on the power data of the abnormal region. Combining the data of non-abnormal regions, the overall power estimation value of the battery pack is recalculated to obtain the corrected overall power status data; S2. Combining the real-time obtained user historical driving records and driving environment parameters, a user behavior feature vector is constructed. According to the user behavior feature vector and the corrected overall power status data, a long short-term memory network model is used to dynamically predict the power consumption trend, and the power consumption rate prediction data under specific scenarios is generated; S3. Combining the current driving environment adaptation parameters, including real-time road conditions and temperature changes, the remaining mileage prediction data is calculated through a preset mileage conversion formula: remaining mileage = (corrected overall power status data / power consumption rate prediction data) * environment adjustment coefficient, where the environment adjustment coefficient is obtained from a preset table based on temperature and road conditions, and finally the preliminary remaining mileage prediction data is obtained; S4. For the preliminary remaining mileage prediction data, a comparative analysis is performed with the recent historical prediction records stored in the system. If the prediction deviation exceeds the preset threshold range, an adaptive correction mechanism is triggered to adjust the parameters of the long short-term memory network model to generate updated model parameters. Combining the latest raw data set of regional power and the user behavior feature vector, the power consumption rate is re-predicted to generate optimized power consumption rate prediction data; S5. For the optimized power consumption rate prediction data, the corrected overall power status data and the driving environment adaptation parameters are continuously updated. Through iterative calculations with a set maximum number of iterations or convergence conditions, the remaining mileage prediction data is dynamically adjusted, and finally the power status and mileage prediction data for real-time display are determined.

2. The method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 1, characterized in that In the step S1, by dividing the battery pack into regions, a multi-point power data acquisition framework is constructed. The power status data of each region is collected in real time in each region to form a raw data set of regional power, including: According to the business attributes of the region division, the battery pack is managed by region, and voltage sensors and current sensors are deployed in each region. The power status of each region is monitored in real time, and the voltage value and current value are obtained from the sensors to obtain the partitioned initial data stream; Through the business attributes of data acquisition, the partitioned initial data stream is transmitted to a pre-established data storage unit, and the voltage values and current values from different regions are integrated and processed by means of data fusion to determine the integrity of the power status of each region and obtain a unified comprehensive data set; If the power status data of a certain area in the comprehensive dataset exceeds the preset threshold range, then for the business attribute of anomaly detection, a preset logic judgment module is called to compare and analyze the voltage value and current value of the area to determine whether there is a potential anomaly; According to the business attribute of the overall power, the power status data of all areas are extracted from the comprehensive dataset. Combining with the business attribute of status evaluation, the data are comprehensively processed by weighted average calculation to obtain the overall power estimation value of the battery pack.

3. A method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 2, characterized in that, In step S1, the voltage and current data of each area are analyzed one by one. If the data of a certain area exceed the preset threshold range, it is marked as an abnormal area, and the abnormal area distribution information is generated, including: According to the pre-established anomaly detection rules, the voltage and current data of each area are obtained from the original sub-area power dataset. For each data point, if the voltage or current of a certain area exceeds the preset threshold range, it is marked as an abnormal area to obtain the preliminary abnormal area distribution information; By performing a secondary verification on the preliminary abnormal area distribution information, the adjacent area data of each abnormal area are obtained. The mean comparison method is used to judge whether the voltage and current of the adjacent areas deviate. If the adjacent area data also exceed the preset threshold range, it is determined that the abnormal area mark is valid, and the confirmed abnormal area list is obtained; According to the confirmed abnormal area list, the time series data of the abnormal areas are extracted from the original dataset. The fluctuation amplitude is calculated for the time series data, and a standard deviation calculation tool is used to judge whether the fluctuation continuously exceeds the preset threshold range to obtain the fluctuation characteristic distribution of the abnormal areas; By classifying the fluctuation characteristic distribution of the abnormal areas, the fluctuation characteristic categories are obtained. For the categories, a data smoothing tool is used for local adjustment, and it is determined whether the adjusted data points return to the preset threshold range to generate the final local data adjustment scheme.

4. A method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 3, characterized in that, In step S1, according to the abnormal area distribution information, a local power correction matrix is constructed to perform weighted adjustment on the power data of the abnormal areas. Combining with the data of the non-abnormal areas, the overall power estimation value of the battery pack is recalculated to obtain the corrected overall power status data, including: According to the abnormal areas and distribution information, a preliminary mapping table for identifying power anomalies is constructed. The power data of each area are obtained from the battery configuration. For the division of the abnormal areas and non-abnormal areas, a preset threshold is used for comparison to obtain the determination result of the specific location and influence range of the abnormal areas; Through the determination result of the abnormal areas, a local power correction matrix is constructed. The power data of the abnormal areas are weighted and adjusted to obtain the adjusted local power values, and they are integrated with the power data of the non-abnormal areas to determine the preliminary overall power distribution table; Using the overall power distribution table and combining with the operating parameters of the battery configuration, the overall power is calibrated for the second time. If the calibrated power estimate does not match the preset threshold range, the correction matrix is finely adjusted to obtain the final overall power status data; Based on the overall power status data, a basic data set for consumption prediction is generated, feature extraction is performed for key indicators, input parameters required for prediction are obtained, it is judged whether the accuracy requirement is met, and the final power estimate is determined.

5. A method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 1, characterized in that, In step S2, by combining the real-time obtained user historical driving records and driving environment parameters, a user behavior feature vector is constructed, including: Acceleration frequency, average vehicle speed and braking frequency data are obtained from the user historical driving records. At the same time, road slope and environmental temperature data are extracted from the driving environment parameters. The data is denoised by a data cleaning tool to obtain the sorted driving behavior data set. A data fusion tool is used to perform correlation mapping on the acceleration frequency and the average vehicle speed and the braking frequency and the road slope to determine the driving habit characteristic values of the user in different environments. ， DH represents the driving habit characteristic value, and T represents the total number of sampling time periods. represents the acceleration at time t, represents the braking force at time t, represents the steering angle at time t, and this formula calculates the comprehensive characteristic value of driving behavior; Through a pre-established calculation rule, the driving habit characteristic values and the environmental temperature data are obtained. If the characteristic values exceed the preset threshold, the power status data is dynamically adjusted to determine the adjusted power consumption baseline value. ， ET represents the environmental temperature impact factor, represents the environmental temperature data, and α, β, and γ are preset temperature impact coefficients. This formula describes the impact relationship of environmental temperature on power consumption; ， δ represents the judgment result of driving habit exceeding the threshold, and DH represents the driving habit eigenvalue. represents the preset driving habit threshold. When the eigenvalue exceeds the threshold, power adjustment is required. ， Represents the adjusted reference value of power consumption, Represents the basic power consumption value, Represents the flag indicating whether adjustment is needed, Represents the driving habit adjustment coefficient, and DH represents the driving habit characteristic value, Represents the temperature influence adjustment coefficient, and ET represents the environmental temperature influence factor. This formula realizes the dynamic adjustment of power consumption according to driving habits and environmental temperature; For the power consumption baseline value, a time series tool is used to perform correlation analysis on the baseline value and the driving mileage data to obtain the user personalized power consumption trend characteristic values.

6. A method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 5, characterized in that, In step S2, based on the user behavior feature vector and the corrected overall power status data, a long short-term memory network model is used to dynamically predict the power consumption trend, and power consumption rate prediction data under specific scenarios is generated, including: Based on the user behavior feature vector and the corrected overall power status data, standardization processing is performed through a pre-established data processing module to obtain the unified feature data set, and outliers are filtered using a preset threshold during processing to obtain the cleaned basic data set; Using the cleaned basic data set, a time series decomposition tool is used to segment the power consumption historical record to obtain the consumption pattern characteristics in different time periods. If the consumption fluctuation in a certain time period exceeds the preset threshold, the data segment is smoothed to determine the intermediate feature set; Through the intermediate feature set, a dynamic model of the power consumption trend is established, and the consumption rate prediction value under the specific scenario is obtained from it. During processing, the feature weights of different scenarios are adjusted to determine the prediction rate data matching the user behavior; Based on the prediction rate data, a scenario adaptation tool is used to combine with the user's current behavior environment information to perform secondary calibration on the prediction value to obtain the final power consumption rate prediction result.

7. A method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 1, characterized in that, Step S3 specifically includes: According to the real-time road condition information and temperature change data, records matching the road conditions and the temperature are queried from a preset environment table to determine the value of the environment adjustment coefficient; By obtaining the corrected power status data, combining with the actual condition of the battery pack power, using an internal calibration tool to calibrate the current available power, and obtaining an accurate corrected power status value; If both the corrected power status value and the predicted power consumption rate data have been obtained, calculate through the mileage conversion formula, that is: remaining mileage = (corrected overall power status data / predicted power consumption rate data) * environmental adjustment factor, where the environmental adjustment factor is obtained from a preset table based on temperature and road conditions. The remaining mileage represents the distance that the vehicle can continue to travel, the corrected overall power status data represents the current available power of the battery pack, the predicted power consumption rate data represents the speed of power consumption per unit time, and the environmental adjustment factor represents the influence factor of the environment on power consumption, to obtain a preliminary remaining mileage prediction result; According to the preliminary remaining mileage prediction result, combining with the driving distance estimation logic, use a data comparison tool to perform a secondary verification on the result, determine whether it meets the reasonable range of consumption per unit time, and determine the final remaining mileage prediction data.

8. A method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 1, characterized in that, In step S4, for the preliminary remaining mileage prediction data, compare and analyze it with the recent historical prediction records stored in the system. If the prediction deviation exceeds the preset threshold range, trigger the adaptive correction mechanism, adjust the parameters of the long short-term memory network model, and generate updated model parameters, including: Obtain the remaining mileage prediction data of the recent record from the repository, compare each item of the recent record with the currently generated prediction data item by item, and obtain the specific value of the prediction deviation by calculating the difference value of each group of data. ， Among them represents the currently generated predicted mileage data represents the actual mileage data of recent records obtained from the repository represents the difference value between the two, that is, the specific value of the prediction deviation; If the prediction deviation exceeds the preset threshold range, trigger the adaptive correction mechanism, determine the parameter adjustment direction and amplitude by comparing the deviation trend in the historical record, and use the pre-established adjustment rules to update the parameters of the long short-term memory network to generate the updated parameter configuration. ， Among them represents a preset threshold range, n represents the number of historical records, represents the prediction deviation value in the historical records, represents the standard deviation of the deviation, represents the adjustment coefficient; ， Among them represents the direction of parameter adjustment, m represents the number of records in the most recent period, represents the prediction deviation; ， wherein represents the final adjustment amplitude, represents the current prediction deviation, represents the average value of historical deviations, and represent the maximum and minimum values of historical deviations respectively, represents the adjustment coefficient; ， Among them represents the updated parameter represents the original parameter represents the direction of parameter adjustment represents the magnitude of adjustment represents the gradient of the parameter The system can update the parameters of the long short-term memory network according to the determined parameter direction and amplitude to generate an updated parameter configuration; determine the change trend of the deviation by comparing the average deviation of the most recent m records with the average deviation of the previous m records, so as to determine the parameter adjustment direction; when the current prediction deviation exceeds this range, the adaptive correction mechanism will be triggered, apply the updated parameter configuration to the prediction process, obtain new remaining mileage prediction data, and compare it with the historical record again to determine whether the prediction deviation has been reduced to within the threshold range.

9. A method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 1, characterized in that, In S4, combine the latest regional power raw data set and the user behavior feature vector to re-predict the power consumption rate and generate optimized power consumption rate prediction data, including: According to the latest regional power raw data set, obtain the power consumption historical records of each region from it, combine with the user behavior feature vector, and use a data cleaning tool to denoise the raw data to obtain a sorted power consumption basic data set; For the sorted power consumption basic data set, use the pre-established user behavior impact weight table to obtain the impact coefficient of each behavior feature on power consumption, and determine the preliminary predicted value of the power consumption rate of each region through a weighted calculation tool. ， Among them represents the preliminary predicted value of the power consumption rate in area i, represents the influence weight coefficient of behavior feature j, represents the value of behavior feature j in area i, and n represents the total number of behavior features; Obtain the preliminary predicted value by weighted summation of each behavior feature. If the deviation between the preliminary predicted value of the power consumption rate and the historical true value exceeds the preset threshold, correct the preliminary predicted value through a regression analysis tool to obtain an adjusted power consumption rate prediction data set. ， wherein represents the predicted deviation percentage of region i, represents the preliminary predicted value of the power consumption rate of region i, represents the historical actual consumption rate of region i, and correction is required when exceeds the preset threshold; ， where CR_i' represents the predicted value of the power consumption rate after correction for region i, represents the preliminary predicted value, and HR_i represents the historical true value, indicating that the range of values of the correction coefficient is from 0 to 1 and is used to control the degree of correction; According to the adjusted power consumption rate prediction data set, combined with the geographical environment data of each region, perform secondary optimization on the prediction data set through a data fusion tool to judge the final power consumption rate prediction result. ， Among them represents the predicted result of the final power consumption rate in area i, represents the corrected predicted value, represents the influence value of the k-th geographical environment factor in area i, represents the weight coefficient of the k-th geographical environment factor, and m represents the total number of geographical environment factors.

10. A method for predicting the state of charge and remaining mileage of an electric vehicle according to claim 1, wherein, The specific steps of step S5 include: According to the power consumption data and the driving environment data, obtain the corresponding adaptation parameters from the pre-established repository. By comparing the current power consumption rate with the historical records, judge whether the adaptation parameters are within the preset threshold range. If they exceed the threshold range, correct the adaptation parameters to obtain an adjusted parameter value. Use the adjusted parameter value, combined with the power status data, and limit the maximum number of iterations and convergence conditions through a cyclic iterative calculation method. Dynamically adjust the remaining mileage prediction data during each iteration to determine a preliminary mileage prediction result. By comparing the preliminary mileage prediction result with the power status data displayed in real time, if the deviation exceeds the preset threshold, re-obtain the latest driving environment data and power consumption data, update the adaptation parameters, and obtain a corrected mileage prediction value. According to the corrected mileage prediction value and the power status data, combined with the real-time display requirements, integrate the final display content through a data fusion method, and judge whether it meets the accuracy requirements. If not, return to the cyclic iterative calculation link to obtain more accurate display data.

Images