Electric vehicle early warning risk grading evaluation method and system based on artificial intelligence

Abstract

The invention relates to the technical field of electric vehicle risk assessment, and particularly discloses an electric vehicle early warning risk grading assessment method and system based on artificial intelligence, and the method comprises the steps: obtaining battery temperature and voltage data during the use of an electric vehicle through a battery management system, carrying out the data processing, and extracting a temperature fluctuation index and a voltage fluctuation index; converting the two fluctuation indexes into a comprehensive feature vector, inputting the comprehensive feature vector into a random forest model, and predicting the risk of the electric vehicle by using the model to obtain a risk coefficient of the electric vehicle; according to the risk coefficients, the risk levels of the electric vehicle are divided into first-level risks, second-level risks and third-level risks, and corresponding early warning measures are taken; through the intelligent risk assessment and real-time monitoring system, the safety of the electric vehicle can be effectively improved, accidents caused by battery faults can be prevented, data support is provided for vehicle maintenance, and the system has important practical application value.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle risk assessment, and particularly to an early warning risk classification and assessment method and system for electric vehicles based on artificial intelligence. Background Art

[0002] With the rapid development of electric vehicle technology, electric vehicles have gradually become an important part of the global transportation industry. As the core component of electric vehicles, the stability of the battery's performance is directly related to the safety, endurance, and service life of the entire vehicle. Since the battery of an electric vehicle may face problems such as high temperature, overcharging, over-discharging, and aging during operation, the state monitoring and fault prediction of the battery have become a key problem in electric vehicle technology.

[0003] To improve the safety and stability of electric vehicles, many studies have begun to explore battery health management and risk prediction methods based on artificial intelligence (AI) technology. By applying machine learning algorithms, especially deep learning and ensemble learning methods, it is possible to comprehensively analyze multi-dimensional data such as the temperature and voltage of the battery, so as to identify potential problems of the battery in advance. However, most of the existing electric vehicle risk prediction systems currently do not incorporate a comprehensive assessment of temperature and voltage fluctuations in the process of battery management, nor do they have a risk classification early warning mechanism. Therefore, how to combine the temperature fluctuations and voltage fluctuations of electric vehicles to construct an efficient and intelligent risk assessment system has become an urgent problem to be solved. Summary of the Invention

[0004] The purpose of the present invention is to provide an early warning risk classification and assessment method and system for electric vehicles based on artificial intelligence to solve the problems in the above background.

[0005] The purpose of the present invention can be achieved by the following technical solutions: An early warning risk classification and assessment method for electric vehicles based on artificial intelligence, comprising the following steps: S1: Obtain the battery temperature data during the use of the electric vehicle, process the temperature data, extract the temperature fluctuation characteristics, and generate a temperature fluctuation index; S2: Obtain the battery voltage data during the use of the electric vehicle, process the voltage data, extract the voltage fluctuation characteristics, and generate a voltage fluctuation index; S3: Convert the temperature fluctuation index and voltage fluctuation index during the use of the electric vehicle into a comprehensive feature vector, use the comprehensive feature vector as the input of the random forest model, and use the random forest model to predict the risks existing in the electric vehicle; S4: According to the prediction results, classify the electric vehicle risk level into primary risk, secondary risk, and tertiary risk; S5: Based on the secondary risk, increase the monitoring frequency, and based on the primary risk, perform early warning processing.

[0006] As a further solution of the present invention: obtaining the battery temperature data during the use of the electric vehicle, processing the temperature data, extracting the temperature fluctuation characteristics, and generating a temperature fluctuation index, specifically including: During the monitoring period, through the temperature sensor in the battery management system, obtain the real-time temperature data of the battery according to the time series; Divide the monitoring period into several identical time windows, calculate the difference between the maximum temperature value and the minimum temperature value in each time window, and obtain the temperature fluctuation amplitude value of each time window; Calculate the standard deviation value of the temperature data in each time window through the standard deviation calculation formula; Normalize the temperature fluctuation amplitude value and the standard deviation value in each time window, calculate the temperature fluctuation index of each time window, and record it as the window temperature fluctuation index; Calculate the average value of the temperature fluctuation indexes in all time windows during the monitoring period to obtain the temperature fluctuation index during the monitoring period.

[0007] As a further solution of the present invention: obtaining the battery voltage data during the use of the electric vehicle, processing the voltage data, extracting the voltage fluctuation characteristics, and generating a voltage fluctuation index, specifically including: During the monitoring period, through the voltage sensor in the battery management system, obtain the real-time voltage data of the battery according to the time series; Divide the monitoring period into several identical time windows, and calculate the standard deviation value and the average value in each time window; Normalize the standard deviation value and the average value in each time window, calculate the voltage fluctuation ratio, and sum up the voltage fluctuation ratios in all windows to obtain the voltage fluctuation index during the monitoring period.

[0008] As a further solution of the present invention: converting the temperature fluctuation index and the voltage fluctuation index during the use of the electric vehicle into a comprehensive feature vector, using the comprehensive feature vector as the input of the random forest model, and predicting the risks existing in the electric vehicle by using the random forest model, specifically including: Obtain the temperature fluctuation index and the voltage fluctuation index of N monitoring periods; For each period, we combine its standardized temperature and voltage fluctuation indexes into a feature vector; Construct the final comprehensive feature vector X through splicing or other feature engineering techniques; wherein, N represents the number of monitoring periods, and N is a positive integer greater than 0; Use the standardized comprehensive feature vector X as the input of the random forest model; The risk coefficient of the electric vehicle is the output of the random forest model, and the model is trained.

[0009] As a further solution of the present invention: using the random forest model to predict the risks existing in the electric vehicle and predicting the risk coefficient of the electric vehicle specifically includes: The random forest model consists of M decision trees; For each decision tree, a subset of the training data is randomly selected, and for each feature, a random subset is selected for splitting to construct the decision tree; For each test sample, all decision trees will give a predicted value of the risk coefficient of the electric vehicle; The final predicted value of the risk coefficient of the electric vehicle is the average of the predicted values of all decision trees.

[0010] As a further solution of the present invention: according to the prediction results, the risk levels of the electric vehicle are divided into first-level risk, second-level risk, and third-level risk, specifically including: Within a real-time monitoring period, obtain the temperature fluctuation index and voltage fluctuation index of the real-time electric vehicle, and convert the temperature fluctuation index and voltage fluctuation index of the real-time electric vehicle into feature vectors as the input of the random forest model; Output the predicted value of the risk coefficient of the electric vehicle; Compare the predicted value of the risk coefficient of the electric vehicle with the first threshold; If the predicted value of the risk coefficient of the electric vehicle is greater than or equal to the first threshold, it is recorded as the first-level risk; If the predicted value of the risk coefficient of the electric vehicle is less than the first threshold, then compare the predicted value of the risk coefficient of the electric vehicle with the second threshold; If the predicted value of the risk coefficient of the electric vehicle is greater than or equal to the second threshold, it is recorded as the second-level risk; If the predicted value of the risk coefficient of the electric vehicle is less than the second threshold, it is recorded as the third-level risk.

[0011] As a further solution of the present invention: based on the second-level risk, increase the monitoring frequency, and based on the first-level risk, perform early warning processing, specifically including: Based on the second-level risk, change the monitoring period to half of the original monitoring period, and collect the time series of voltage data and temperature data according to half of the original time series; Based on the first-level risk, trigger the early warning mechanism and immediately stop the use of the electric vehicle; The system can remotely control to restrict the owner from continuing to use the vehicle until the risk is handled; After troubleshooting, conduct a comprehensive inspection of the vehicle, and after confirming its safety, re-evaluate whether it can continue to be used.

[0012] An artificial intelligence-based early warning risk classification and assessment system for electric vehicles, comprising: a data acquisition module for acquiring battery temperature data and battery voltage data when the electric vehicle is in use; A data processing module for processing the temperature data, extracting temperature fluctuation characteristics, generating a temperature fluctuation index, processing the voltage data, extracting voltage fluctuation characteristics, and generating a voltage fluctuation index; A risk prediction module that converts the temperature fluctuation index and voltage fluctuation index during the use of the electric vehicle into a comprehensive feature vector, uses the comprehensive feature vector as the input of a random forest model, and uses the random forest model to predict the risks existing in the electric vehicle; A risk level classification module that classifies the risk levels of electric vehicles into primary risks, secondary risks, and tertiary risks according to the prediction results; An early warning processing module that increases the monitoring frequency based on secondary risks and performs early warning processing based on primary risks.

[0013] For the artificial intelligence-based early warning risk classification and assessment system for electric vehicles, the data processing module further includes at least: A temperature fluctuation index calculation unit for calculating the temperature fluctuation index by acquiring the temperature data within the monitoring period; A voltage fluctuation index calculation unit for calculating the voltage fluctuation index by acquiring the voltage data within the monitoring period.

[0014] For the artificial intelligence-based early warning risk classification and assessment system for electric vehicles, the risk prediction module further includes at least: A risk coefficient acquisition unit for obtaining the risk coefficient of the electric vehicle within the monitoring period by processing the temperature fluctuation index and voltage fluctuation index.

[0015] Advantages of the present invention: (1) By real-time monitoring and analyzing the temperature and voltage fluctuations of the electric vehicle battery, potential risk factors can be accurately identified; by extracting the temperature fluctuation index and voltage fluctuation index, the system can comprehensively evaluate the working state and stability of the battery, thereby predicting possible faults or abnormal conditions of the battery; when abnormal fluctuations occur in the battery temperature or voltage, the system will react in a timely manner, identify potential risks through in-depth analysis of the data; by classifying risks into first-level, second-level, and third-level categories, the system can take different countermeasures for different risk levels. For example, when there is a first-level risk, the vehicle is immediately stopped from being used, and when there is a second-level risk, the monitoring frequency is increased and a health report is generated; this precise early warning mechanism can effectively prevent fires, faults, or other safety hazards caused by battery problems, thereby improving the overall safety and reliability of electric vehicles, reducing the accident rate caused by battery failures, and enhancing the driving confidence of vehicle owners; (2) Through the electric vehicle early warning risk assessment system provided by the present invention, vehicle owners and management can achieve intelligent management of the health status of electric vehicles and optimize maintenance and repair strategies; when the vehicle is in a second-level risk state, the system will automatically increase the monitoring frequency, collect and analyze the temperature and voltage of the battery more frequently, and generate a vehicle health status report; these reports can not only help vehicle owners discover potential problems in a timely manner and avoid using damaged vehicles for a long time, but also provide data-based decision-making basis for management, enabling them to arrange maintenance or component replacement in advance; in the first-level risk state, the system will notify vehicle owners and service centers through various communication methods and take emergency measures in a timely manner to avoid more serious accidents; by real-time monitoring and precise analysis of the health status of each electric vehicle. Description of the Drawings

[0016] The present invention will be further described below in conjunction with the accompanying drawings.

[0017] Figure 1 It is a specific step flow block diagram of the electric vehicle early warning risk classification and assessment method based on artificial intelligence of the present invention; Figure 2 It is a flow block diagram of the electric vehicle early warning risk classification and assessment system based on artificial intelligence in the present invention. Detailed Embodiment

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in 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. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0019] Please refer to Figure 1As shown in the figure, the present invention is an early warning risk grading assessment method for electric vehicles based on artificial intelligence, including the following steps: S1: Obtain the battery temperature data during the use of the electric vehicle, process the temperature data, extract the temperature fluctuation characteristics, and generate a temperature fluctuation index; S2: Obtain the battery voltage data during the use of the electric vehicle, process the voltage data, extract the voltage fluctuation characteristics, and generate a voltage fluctuation index; S3: Convert the temperature fluctuation index and voltage fluctuation index during the use of the electric vehicle into a comprehensive feature vector, use the comprehensive feature vector as the input of the random forest model, and use the random forest model to predict the risks existing in the electric vehicle; S4: According to the prediction results, classify the risk levels of electric vehicles into first-level risks, second-level risks, and third-level risks; S5: Based on the second-level risk, increase the monitoring frequency, and based on the first-level risk, conduct early warning processing.

[0020] In S1, obtaining the battery temperature data during the use of the electric vehicle, processing the temperature data, extracting the temperature fluctuation characteristics, and generating a temperature fluctuation index specifically includes: During the monitoring period, through the temperature sensor in the battery management system, obtain the real-time temperature data of the battery according to the time series (such as every minute); Calibrate the temperature sensor to ensure the accuracy of the temperature data; The temperature data may have noise (such as due to sensor errors or external environmental interference), so it is necessary to denoise the collected voltage data, and filtering algorithms (such as low-pass filters or median filters) can be used to eliminate this noise; If data is missing at certain time points, interpolation methods or filling strategies (such as linear interpolation, mean filling) can be used to restore the missing temperature values; Divide the monitoring period into several identical time windows (such as every 10 minutes), calculate the difference between the maximum temperature value and the minimum temperature value within each time window to obtain the temperature fluctuation amplitude value of each time window, denoted as ; Among them, represents the number of time windows, = 1, 2,......, n; Calculate the standard deviation value of the temperature data of each time window through the standard deviation calculation formula ; Normalize the temperature fluctuation amplitude value and standard deviation value within each time window, and calculate the temperature fluctuation index of each time window, denoted as the window temperature fluctuation index; Calculate the average value of the temperature fluctuation indices within all time windows during the monitoring period to obtain the temperature fluctuation index during the monitoring period, denoted as ; wherein, the temperature fluctuation index during the monitoring period has the following calculation expression: ; In the formula, represents the temperature fluctuation index during the monitoring period, and are preset proportionality coefficients, and and are both greater than 0, represents the temperature fluctuation amplitude value within the th time window, represents the standard deviation value within the th time window, represents the number of time windows, and n represents the maximum number of time windows; It should be noted that: the temperature fluctuation index during the monitoring period reflects the fluctuation situation of the temperature data during the monitoring period, and based on the fluctuation situation, it is judged whether the temperature data is abnormal when the electric vehicle is in use. Abnormal temperature data may pose a risk of malfunction to the battery performance.

[0021] In S2, obtain the battery voltage data when the electric vehicle is in use, process the voltage data, extract the voltage fluctuation characteristics, and generate a voltage fluctuation index, which specifically includes: During the monitoring period, obtain the real-time voltage data of the battery through the voltage sensor in the battery management system according to the time series (such as every minute); Calibrate the voltage sensor to ensure the accuracy of the voltage data; The voltage data may contain noise (such as due to sensor errors or external environmental interference), so it is necessary to denoise the collected voltage data. Filtering algorithms (such as low-pass filters or median filters) can be used to eliminate this noise; If data is missing at some time points, interpolation methods or filling strategies (such as linear interpolation, mean filling) can be used to restore the missing voltage values; Divide the monitoring period into several identical time windows (such as every 10 minutes), and calculate the standard deviation and the average value within each time window; Normalize the standard deviation and the average value within each time window, calculate the voltage fluctuation ratio, and sum up the voltage fluctuation ratios within all windows to obtain the voltage fluctuation index during the monitoring period; Among them, the calculation expression of the voltage fluctuation ratio is as follows: ; In the formula, represents the number of time windows, represents the voltage fluctuation ratio within the -th time window, and are preset proportionality coefficients, and and are both greater than 0, represents the standard deviation of the voltage within the -th time window, represents the average voltage within the -th time window; It should be noted that: The voltage fluctuation index of the battery during the use of an electric vehicle reflects the change rate, stability, and abnormal fluctuation of the battery voltage. Moreover, the greater the voltage fluctuation index, the more unstable the corresponding battery voltage.

[0022] In S3, the temperature fluctuation index and voltage fluctuation index during the use of an electric vehicle are converted into a comprehensive feature vector, and the comprehensive feature vector is used as the input of the random forest model to predict the risks existing in the electric vehicle, specifically including: Obtain the temperature fluctuation index and voltage fluctuation index within multiple historical identical monitoring periods; Perform standardization processing on the temperature fluctuation index and voltage fluctuation index within multiple monitoring periods, including standardizing the values and standard deviations to ensure that the data is modeled on the same scale; Obtain the temperature fluctuation index and voltage fluctuation index of N monitoring periods; For each period, we combine its standardized temperature and voltage fluctuation indices into a feature vector; Through concatenation or other feature engineering techniques, we can construct the final comprehensive feature vector X; Among them, X is a 2N-dimensional vector, N represents the number of monitoring periods, and N is a positive integer greater than 0; Compress the data of multiple monitoring periods into a unified feature representation, and use the standardized comprehensive feature vector X as the input of the random forest model; The risk coefficient of the electric vehicle is the output of the random forest model. Train the model and evaluate the model performance using methods such as cross-validation; The random forest model consists of M decision trees; For each decision tree, randomly select a subset of the training data, and for each feature, select a random subset for splitting to construct the decision tree; For each test sample, all decision trees will give a predicted value of the risk coefficient of an electric vehicle; The final predicted value of the risk coefficient of the electric vehicle is the average of the predicted values of all decision trees.

[0023] In S4, according to the prediction results, the risk levels of electric vehicles are divided into first-level risk, second-level risk, and third-level risk, which specifically include: Among them, the first-level risk is a severe risk level, the second-level risk is a general risk level, and the third-level risk means no risk; During a real-time monitoring period, obtain the temperature fluctuation index and voltage fluctuation index of the real-time electric vehicle, and convert the temperature fluctuation index and voltage fluctuation index of the real-time electric vehicle into feature vectors as the input of the random forest model; Output the predicted value of the risk coefficient of the electric vehicle; Compare the predicted value of the risk coefficient of the electric vehicle with the first threshold; If the predicted value of the risk coefficient of the electric vehicle is greater than or equal to the first threshold, it indicates that the risk level during the use of the corresponding electric vehicle is very high and the probability of failure is high, which is recorded as the first-level risk; If the predicted value of the risk coefficient of the electric vehicle is less than the first threshold, then compare the predicted value of the risk coefficient of the electric vehicle with the second threshold; If the predicted value of the risk coefficient of the electric vehicle is greater than or equal to the second threshold, it indicates that the risk level during the use of the corresponding electric vehicle is general and the probability of failure is low, which is recorded as the second-level risk; If the predicted value of the risk coefficient of the electric vehicle is less than the second threshold, it indicates that the probability of risk during the use of the corresponding electric vehicle is close to 0, which is recorded as the third-level risk; In S5, based on the second-level risk, increase the monitoring frequency, and based on the first-level risk, conduct early warning processing, which specifically includes: Based on the second-level risk, change the monitoring period to half of the original monitoring period, and collect the time series of voltage data and temperature data according to half of the original time series; And generate a vehicle health status report every day, which contains all abnormal data points and their analysis conclusions for the reference of management decision-making; Open a dedicated user communication channel. When the vehicle is at the second-level risk, inform the vehicle owner of the vehicle detection status through methods such as APP push and text messages, and avoid long-term use; Based on the first-level risk, immediately notify the vehicle owner and the nearest service center through various methods (such as phone calls, text messages, APP messages), inform the risks existing in the electric vehicle, and immediately stop using it; The service center should receive the notice in a timely manner, prepare treatment measures, and arrange technicians to conduct fault troubleshooting and repair; To ensure the safety of vehicle owners, in the case of first-level risks, the vehicle should immediately stop being used; The system can turn off some non-critical functions through remote control to restrict the vehicle owner from continuing to use the vehicle until the risk is handled; After the problem is solved, conduct in-depth fault analysis to identify the root cause of the problem, and improve the system through data analysis to reduce the occurrence of future risks; After troubleshooting, conduct a comprehensive inspection of the vehicle, and after confirming its safety, re-evaluate whether it can be put into use.

[0024] Please refer to Figure 2 As shown, the artificial intelligence-based early warning risk classification and assessment system for electric vehicles includes: A data acquisition module, which is used to obtain battery temperature data and battery voltage data when the electric vehicle is in use; A data processing module, which is used to process the temperature data, extract temperature fluctuation characteristics, generate a temperature fluctuation index, process the voltage data, extract voltage fluctuation characteristics, and generate a voltage fluctuation index; A risk prediction module, which converts the temperature fluctuation index and voltage fluctuation index when the electric vehicle is in use into a comprehensive feature vector, uses the comprehensive feature vector as the input of the random forest model, and uses the random forest model to predict the risks existing in the electric vehicle; A risk level classification module, which classifies the electric vehicle risk level into first-level risk, second-level risk, and third-level risk according to the prediction results; An early warning processing module, which increases the monitoring frequency based on the second-level risk and conducts early warning processing based on the first-level risk.

[0025] Working principle of the present invention: By using the temperature sensor and voltage sensor in the battery management system, the temperature and voltage data of the electric vehicle are collected, and these data are processed by the data processing module, such as denoising, calibration, and interpolation, to extract the temperature fluctuation characteristics and voltage fluctuation characteristics, and generate the temperature fluctuation index and voltage fluctuation index respectively. Subsequently, the temperature fluctuation index and voltage fluctuation index are converted into a comprehensive feature vector, which is used as input data to be fed into the random forest model to predict the risk coefficient of the electric vehicle. According to the prediction result of the risk coefficient, the risk level classification module classifies the risk level of the electric vehicle into first-level risk, second-level risk, and third-level risk, where the first-level risk represents high risk, the second-level risk represents general risk, and the third-level risk represents low risk. For the second-level risk, the system will automatically increase the monitoring frequency, shorten the monitoring period, and generate a vehicle health status report to remind the management to take measures; for the first-level risk, the system will notify the vehicle owner and the service center through various methods such as phone calls, text messages, and APP messages, and restrict the use of the vehicle through the remote control system to prevent safety accidents from occurring. In addition, the early warning processing module will take corresponding measures according to the risk level, such as strengthening monitoring and early warning in the case of second-level risk, and immediately stopping the vehicle use and performing maintenance processing in the case of first-level risk. The core advantage of the entire system lies in the accurate prediction and risk assessment of battery temperature and voltage fluctuations through artificial intelligence algorithms, timely discovery of potential faults and dynamic adjustment, thereby effectively improving the safety and reliability of electric vehicles.

[0026] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0027] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0028] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood with reference to the context before and after.

[0029] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0030] The above has described in detail one embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the present application should still fall within the scope covered by the patent of the present invention.

Claims

1. An artificial intelligence-based early warning risk grading and assessment method for electric vehicles, characterized in that, Including the following steps: S1: Obtain the battery temperature data when the electric vehicle is in use, process the temperature data, extract the temperature fluctuation characteristics, and generate a temperature fluctuation index; S2: Obtain the battery voltage data when the electric vehicle is in use, process the voltage data, extract the voltage fluctuation characteristics, and generate a voltage fluctuation index; S3: Convert the temperature fluctuation index and voltage fluctuation index when the electric vehicle is in use into a comprehensive feature vector, use the comprehensive feature vector as the input of the random forest model, and use the random forest model to predict the risks existing in the electric vehicle; S4: According to the prediction results, divide the risk levels of the electric vehicle into first-level risks, second-level risks, and third-level risks; S5: Based on the second-level risks, increase the monitoring frequency, and based on the first-level risks, conduct early warning processing.

2. The method for warning risk classification and assessment of an electric vehicle based on artificial intelligence according to claim 1, wherein, The obtaining of the battery temperature data when the electric vehicle is in use, processing the temperature data, extracting the temperature fluctuation characteristics, and generating a temperature fluctuation index specifically includes: During the monitoring period, obtain the real-time temperature data of the battery according to the time series through the temperature sensor in the battery management system; Divide the monitoring period into several identical time windows, calculate the difference between the maximum temperature value and the minimum temperature value in each time window, and obtain the temperature fluctuation amplitude value of each time window; Calculate the standard deviation value of the temperature data in each time window through the standard deviation calculation formula; Normalize the temperature fluctuation amplitude value and the standard deviation value in each time window, calculate the temperature fluctuation index of each time window, and record it as the window temperature fluctuation index; Calculate the average value of the temperature fluctuation indexes in all time windows during the monitoring period to obtain the temperature fluctuation index during the monitoring period.

3. The method for grading and evaluating the early warning risk of an electric vehicle based on artificial intelligence according to claim 1, wherein The obtaining of the battery voltage data when the electric vehicle is in use, processing the voltage data, extracting the voltage fluctuation characteristics, and generating a voltage fluctuation index specifically includes: During the monitoring period, obtain the real-time voltage data of the battery according to the time series through the voltage sensor in the battery management system; Divide the monitoring period into several identical time windows, and calculate the standard deviation value and the average value in each time window; Normalize the standard deviation value and the average value in each time window, calculate the voltage fluctuation ratio, and sum up the voltage fluctuation ratios in all windows to obtain the voltage fluctuation index during the monitoring period.

4. The method for grading and evaluating the early warning risks of an electric vehicle based on artificial intelligence according to claim 1, wherein The conversion of the temperature fluctuation index and voltage fluctuation index when the electric vehicle is in use into a comprehensive feature vector, using the comprehensive feature vector as the input of the random forest model, and using the random forest model to predict the risks existing in the electric vehicle specifically includes: Obtain the temperature fluctuation index and voltage fluctuation index of N monitoring periods; For each period, we combine its standardized temperature and voltage fluctuation indexes into a feature vector; Construct the final comprehensive feature vector X through concatenation or other feature engineering techniques; Among them, N represents the number of monitoring periods, and N is a positive integer greater than 0; Use the standardized comprehensive feature vector X as the input of the random forest model; The risk coefficient of the electric vehicle is the output of the random forest model, and the model is trained.

5. The method for grading and evaluating the warning risk of an electric vehicle based on artificial intelligence according to claim 4, wherein The use of the random forest model to predict the risks existing in the electric vehicle and predict the risk coefficient of the electric vehicle specifically includes: The random forest model consists of M decision trees; For each decision tree, a subset of the training data is randomly selected, and for each feature, a random subset is selected for splitting to construct the decision tree; For each test sample, all decision trees will give a predicted value of the risk coefficient of the electric vehicle; The final predicted value of the risk coefficient of the electric vehicle is the average of the predicted values of all decision trees.

6. The method for grading and evaluating the warning risk of an electric vehicle based on artificial intelligence according to claim 1, wherein, According to the prediction results, the risk levels of electric vehicles are divided into first-level risk, second-level risk, and third-level risk, specifically including: During a real-time monitoring period, obtain the temperature fluctuation index and voltage fluctuation index of the real-time electric vehicle, and convert the temperature fluctuation index and voltage fluctuation index of the real-time electric vehicle into feature vectors as the input of the random forest model; Output the predicted value of the risk coefficient of the electric vehicle; Compare the predicted value of the risk coefficient of the electric vehicle with the first threshold; If the predicted value of the risk coefficient of the electric vehicle is greater than or equal to the first threshold, it is recorded as a first-level risk; If the predicted value of the risk coefficient of the electric vehicle is less than the first threshold, then compare the predicted value of the risk coefficient of the electric vehicle with the second threshold; If the predicted value of the risk coefficient of the electric vehicle is greater than or equal to the second threshold, it is recorded as a second-level risk; If the predicted value of the risk coefficient of the electric vehicle is less than the second threshold, it is recorded as a third-level risk.

7. The method for grading and evaluating the early warning risk of an electric vehicle based on artificial intelligence according to claim 1, wherein, Based on the second-level risk, increase the monitoring frequency, and based on the first-level risk, conduct early warning processing, specifically including: Based on the second-level risk, change the monitoring period to half of the original monitoring period, and collect the time series of voltage data and temperature data according to half of the original time series; Based on the first-level risk, trigger the early warning mechanism and immediately stop the use of the electric vehicle; The system can remotely control to restrict the owner from continuing to use the vehicle until the risk is handled; After troubleshooting, conduct a comprehensive inspection of the vehicle, and after confirming its safety, re-evaluate whether it can continue to be used.

8. An early warning risk grading and assessment system for electric vehicles based on artificial intelligence, characterized in that, For the artificial intelligence-based electric vehicle early warning risk classification and evaluation method as described in any one of claims 1-7, including: a data acquisition module, the data acquisition module is used to obtain the battery temperature data and battery voltage data when the electric vehicle is in use; A data processing module, the data processing module is used to process the temperature data, extract the temperature fluctuation characteristics, generate the temperature fluctuation index, process the voltage data, extract the voltage fluctuation characteristics, and generate the voltage fluctuation index; A risk prediction module, the risk prediction module converts the temperature fluctuation index and voltage fluctuation index when the electric vehicle is in use into a comprehensive feature vector, uses the comprehensive feature vector as the input of the random forest model, and uses the random forest model to predict the risks existing in the electric vehicle; A risk level classification module, the risk level classification module divides the risk levels of electric vehicles into first-level risk, second-level risk, and third-level risk according to the prediction results; An early warning processing module, the early warning processing module increases the monitoring frequency based on the second-level risk and conducts early warning processing based on the first-level risk.

9. The artificial intelligence-based early warning risk classification and assessment system for electric vehicles according to claim 8, wherein The data processing module at least further includes: A temperature fluctuation index calculation unit, the temperature fluctuation index calculation unit calculates the temperature fluctuation index by obtaining the temperature data within the monitoring period; A voltage fluctuation index calculation unit, which calculates the voltage fluctuation index by obtaining voltage data within a monitoring period.

10. The artificial intelligence-based electric vehicle warning risk classification and evaluation system according to claim 8, wherein The risk prediction module further includes at least: A risk coefficient acquisition unit, which obtains the risk coefficient within the monitoring period of the electric vehicle by processing the temperature fluctuation index and the voltage fluctuation index.

Images