Battery health state evaluation method and equipment based on vehicle driving data

Abstract

The invention discloses a battery health state evaluation method and device based on vehicle driving data, and relates to the technical field of electric vehicle big data analysis, and the method comprises the steps: determining a model input vector from two aspects of battery health features and feature influence factor extraction; on one hand, capacity consumption and a current peak value are calculated, and an empirical mode decomposition method is used for assisting in extracting battery health features. On the other hand, a residual electric quantity sequence and an average temperature sequence are determined, and a Pearson's correlation coefficient method is used for analysis to obtain characteristic influence factors. And finally, constructing an integrated learning model by using the model input vector to realize evaluation of the battery health state, thereby effectively solving the problem of excessive dependence on charging data in existing power battery health state estimation, improving the utilization efficiency of original signals, realizing accurate estimation of the battery health state, and improving the reliability of the power battery health state estimation. And the consumption of manpower, material resources and financial resources caused by the design calibration experiment of the power battery of the electric vehicle and the re-measurement of data is avoided.

Description

Technical Field

[0001] The present application relates to the technical field of electric vehicle big data analysis, and in particular to a battery health status assessment method and device based on vehicle driving data. Background Art

[0002] The development of electric vehicles is one of the key means to solve the fossil energy crisis, alleviate global warming, and promote carbon neutrality. As the core component of electric vehicles, the state of health (SOH) of power batteries has an important impact on the driving performance and safety of the vehicle. Therefore, accurate estimation of power battery SOH is one of the key guarantees for the safe and stable operation of electric vehicles. Traditional battery SOH evaluation methods based on electric vehicle big data mostly rely on relatively stable electric vehicle charging data, ignoring electric vehicle driving data with richer data resources, resulting in low utilization of real vehicle data and affecting the accuracy of the evaluation. Summary of the invention

[0003] Compared with traditional fuel vehicles, electric vehicles have significant advantages in on-board data collection and processing. If the battery health factor can be extracted based on real vehicle driving data, the utilization rate of real vehicle big data can be improved, the reliance of existing technologies on vehicle charging data can be reduced, and the battery aging mechanism contained in the changes in vehicle power performance can be fully considered. Therefore, the purpose of this application is to provide a battery health status assessment method and device based on vehicle driving data, which can improve the utilization rate of real vehicle data and achieve accurate assessment of battery health status.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a battery health status assessment method based on vehicle driving data, comprising:

[0006] Acquire historical driving data of the electric vehicle and retain corresponding reference data items; the reference data items include: vehicle driving speed, cumulative mileage, current, remaining power, latitude and longitude coordinates, maximum temperature, minimum temperature and vehicle identification code;

[0007] Selecting a target segment from the historical driving data according to the vehicle identification code, latitude and longitude coordinates and vehicle driving speed in the reference data item;

[0008] Calculating a battery health status reference value according to the cumulative mileage, maximum temperature and minimum temperature of the target segment;

[0009] Calculating capacity consumption and current peak value according to the current of the target segment;

[0010] Performing empirical mode decomposition on the capacity consumption and current peak of the target segment to obtain components of the capacity consumption and components of the current peak, and using the Pearson correlation coefficient to select components whose correlation with the battery health state reference value is within a set range as battery health features;

[0011] Determine a remaining power sequence and an average temperature sequence according to the remaining power, maximum temperature, and minimum temperature of the target segment, and use the Pearson correlation coefficient to select a sequence whose correlation with the battery health feature is within a set range as a feature influencing factor;

[0012] Constructing an integrated learning model, and training the integrated learning model according to the battery health characteristics, characteristic influencing factors and corresponding battery health status reference values ​​of the target segment to obtain a trained integrated learning model;

[0013] The trained integrated learning model is used to evaluate the battery health state based on the current driving data of the electric vehicle to obtain a battery health state prediction value.

[0014] Optionally, historical driving data of the electric vehicle is obtained, and corresponding reference data items are retained, including:

[0015] From the actual vehicle data of electric vehicles collected on the big data platform, the historical driving data of electric vehicles is screened based on the vehicle driving speed not being equal to 0, and the corresponding reference data items are retained.

[0016] Optionally, selecting a target segment from the historical driving data according to the vehicle identification code, latitude and longitude coordinates and vehicle driving speed in the reference data item includes:

[0017] According to the vehicle identification code and the latitude and longitude coordinates, a data set having the same vehicle model and vehicle driving route is selected from the historical driving data;

[0018] According to the longitude and latitude coordinates, a straight line track having a mileage span less than or equal to a set distance is selected from the data set for segmentation to obtain a plurality of discharge segments;

[0019] The average vehicle speed of each of the discharge segments is calculated according to the vehicle running speed, the average vehicle speed interval is divided, and the discharge segment with the largest average vehicle speed interval proportion is selected as the target segment.

[0020] Optionally, calculating a battery health state reference value according to the cumulative mileage, maximum temperature, and minimum temperature of the target segment includes:

[0021] The battery health status calculation formula is SOH = 1.084M -0.01884 -0.04368+0.00572T m+1.657×10 - 7 M m -0.0001524T m 2 +6.964×10 -9 T m M m , calculate the battery health state reference value SOH according to the cumulative mileage, maximum temperature and minimum temperature of the target segment; where M represents the cumulative mileage of the target segment, T m represents the monthly average temperature calculated based on the maximum and minimum temperatures of the target segment, M m Indicates the monthly cumulative mileage calculated based on the cumulative mileage of the target segment.

[0022] Optionally, calculating the capacity consumption and the current peak value according to the current of the target segment includes:

[0023] Using the capacity consumption calculation formula Calculate the capacity consumption CC according to the current of the target segment; wherein I(i) represents the current of the i-th sampling point in the target segment, and n represents the number of sampling points in the target segment;

[0024] The current peak value is calculated using the formula I max =max(I(1), I(2), ..., I(n)), calculate the current peak value I according to the current of the target segment max ; Wherein, (I(1), I(2), ..., I(n)) represents the current of all sampling points in the target segment.

[0025] Optionally, performing empirical mode decomposition on the capacity consumption and the current peak of the target segment to obtain components of the capacity consumption and components of the current peak includes:

[0026] Performing empirical mode decomposition on the capacity consumption of the target segment to obtain components of the capacity consumption; the components of the capacity consumption include: a plurality of intrinsic mode functions corresponding to the capacity consumption and a residual;

[0027] Performing empirical mode decomposition on the current peak of the target segment to obtain components of the current peak; the components of the current peak include: a plurality of intrinsic mode functions corresponding to the current peak and a residual.

[0028] Optionally, determining a remaining power sequence and an average temperature sequence according to the remaining power, the maximum temperature, and the minimum temperature of the target segment includes:

[0029] Extracting the remaining power of the target segment to obtain a remaining power sequence;

[0030] Extract the maximum and minimum temperatures of the target fragment and use the formula and Calculate the average temperature T of the target fragment a , and get the average temperature series; where T i represents the average temperature of the i-th sampling point in the target segment, n represents the number of sampling points in the target segment, T imax represents the maximum temperature of the i-th sampling point in the target segment, T imin Indicates the minimum temperature of the i-th sampling point in the target segment.

[0031] Optionally, the basic models of the integrated learning model include: two Gaussian process regression models and two support vector machine models; the output of the integrated learning model is the weighted sum of the outputs of the four basic models; during the training process, the weights of the four basic models are updated once every 5,000 kilometers, every 10,000 kilometers or every 20,000 kilometers according to the error of each basic model.

[0032] Optionally, the battery health status assessment method based on vehicle driving data further includes:

[0033] A battery degradation trajectory is plotted based on the battery health status prediction value.

[0034] In a second aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the battery health status assessment method based on vehicle driving data.

[0035] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0036] The present application provides a battery health status assessment method and device based on vehicle driving data, by obtaining the historical driving data of electric vehicles and retaining the corresponding reference data items, namely, vehicle driving speed, cumulative mileage, current, remaining power, latitude and longitude coordinates, maximum temperature, minimum temperature and vehicle identification code, the battery health status reference value, capacity consumption, current peak, remaining power sequence and average temperature sequence are determined based on the reference data items of the historical driving data, and the empirical mode decomposition and Pearson correlation coefficient are used to assist in extracting the battery health characteristics and characteristic influencing factors during driving, and an integrated learning model is constructed and trained. The battery health degradation mechanism related to driving behavior, operating conditions, etc. contained in the driving data is used to improve the accuracy of the battery health status assessment from the dimension of vehicle power performance changes. Therefore, the present application can improve the utilization rate of real vehicle data and realize accurate assessment of battery health status. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 A flowchart of a battery health status assessment method based on vehicle driving data provided in this application;

[0039] Figure 2 A graph showing the proportion of different average speed ranges for a certain type of vehicle provided for this application;

[0040] Figure 3 Empirical modal decomposition of capacity consumption provided for this application;

[0041] Figure 4 Empirical mode decomposition diagram of current peak provided for this application;

[0042] Figure 5 Framework diagram of the integrated learning model provided for this application. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0044] Due to their high degree of electrification, electric vehicles will generate a large amount of real vehicle data during operation. Their driving data contains a large number of battery health degradation mechanisms related to driving behavior, operating conditions, etc., which is conducive to increasing the accuracy of battery SOH evaluation from the dimension of vehicle power performance changes. The organic combination of real vehicle operation data and battery health degradation mechanism will help to achieve accurate estimation of electric vehicle power battery SOH. In view of this, the present application provides a battery health status evaluation method and device based on vehicle driving data, which can improve the utilization rate of real vehicle data and achieve accurate evaluation of battery health status.

[0045] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0046] In an exemplary embodiment, the present application provides a battery health status assessment method based on vehicle driving data, which is executed by a computer device, and can be specifically executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, Figure 1 As shown, the method includes the following steps 1 to 8.

[0047] Step 1: Obtain the historical driving data of the electric vehicle and retain the corresponding reference data items.

[0048] Specifically, from the actual vehicle data of electric vehicles collected on the big data platform, the historical driving data of electric vehicles is screened based on the vehicle driving speed not being equal to 0, and the corresponding reference data items are retained.

[0049] The reference data items include: vehicle speed, accumulated mileage, current, remaining power, longitude and latitude coordinates, maximum temperature, minimum temperature and vehicle identification code; the longitude and latitude coordinates include: longitude and latitude.

[0050] Step 2: Select a target segment from the historical driving data according to the vehicle identification code, longitude and latitude coordinates and vehicle driving speed in the reference data item.

[0051] Step 2.1: According to the vehicle identification code and the latitude and longitude coordinates, a data set with the same vehicle model and vehicle driving route is selected from the historical driving data.

[0052] Step 2.2: According to the latitude and longitude coordinates, select straight line tracks with a mileage span less than or equal to a set distance from the data set for segmentation to obtain multiple discharge segments. The set distance is preferably 2 kilometers.

[0053] Step 2.3: Calculate the average vehicle speed of each of the discharge segments according to the vehicle running speed, divide the average vehicle speed interval, and select the discharge segment with the largest average vehicle speed interval as the target segment.

[0054] The calculation formula of the average vehicle speed is:

[0055]

[0056] in, represents the average vehicle speed in the discharge segment, V i represents the vehicle speed at the i-th sampling point in the discharge segment, and n represents the number of sampling points in the discharge segment (which is also equal to the number of sampling points in the target segment).

[0057] like Figure 2As shown, the average speed of a certain model of vehicle can be divided into three intervals, namely less than 20km / h, 20-25km / h and greater than 25km / h. If an interval accounts for more than 1 / 3, it will be retained. If multiple intervals are retained in the previous screening step, the interval with the largest proportion, that is, the discharge segment in the 20-25km / h interval, is selected as the target segment.

[0058] In this application, by selecting a straight track, excessive fluctuations in the data caused by oversteering of the vehicle can be avoided. By limiting the length of the straight track and selecting a shorter segment, the span of the data can be reduced, avoiding fluctuations in the data caused by long vehicle driving time and many unstable factors. In addition, by selecting segments with a large proportion of speed intervals for retention, the richness of the data volume can be guaranteed as much as possible.

[0059] Step 3: Calculate a battery health status reference value according to the cumulative mileage, maximum temperature and minimum temperature of the target segment.

[0060] The calculation formula of the battery health status reference value is:

[0061] SOH=1.084M -0.01884 -0.04368+0.00572T m +1.657×10 -7 M m -0.0001524T m 2 +6.964×10 -9 T m M m

[0062] Among them, SOH represents the reference value of battery health status, M represents the cumulative mileage of the target segment, and T m represents the monthly average temperature calculated based on the maximum and minimum temperatures of the target segment, M m Indicates the monthly cumulative mileage calculated based on the cumulative mileage of the target segment.

[0063] The calculation formula for the monthly average temperature is:

[0064]

[0065]

[0066] in, Indicates the average temperature on the day of this month, T Times represents the average temperature of the Times sampling point in the day, k represents the number of sampling points in the day, and D represents the number of days in the month.

[0067] The monthly accumulated mileage is calculated as follows:

[0068] M m =M last -M first

[0069] Among them, M last Indicates the accumulated mileage at 24:00 on the last day of the month, M first Indicates the accumulated mileage at 0:00 on the first day of the month.

[0070] Step 4: Calculate the capacity consumption (CC) and the maximum current (I) according to the current of the target segment. max ), and further draw the evolution trajectory diagram.

[0071] The capacity consumption evolution trajectory diagram has the accumulated mileage as the horizontal axis and the capacity consumption as the vertical axis, see Figure 3 The first column and the first figure in the figure; the evolution trajectory diagram of the current peak value has the accumulated mileage as the horizontal axis and the current peak as the vertical axis, see Figure 4 The first picture in the first column.

[0072] The calculation formula of the capacity consumption is:

[0073]

[0074] Wherein, CC represents the capacity consumption of the target segment, I(i) represents the current of the i-th sampling point in the target segment, and n represents the number of sampling points in the target segment.

[0075] The calculation formula of the current peak value is:

[0076] I max =max(I(1),I(2),...,I(n))

[0077] Among them, I max represents the peak current of the target segment, and (I(1), I(2), ..., I(n)) represents the current of all sampling points in the target segment.

[0078] Step 5: Perform empirical mode decomposition (EMD) on the capacity consumption and current peak of the training segment to obtain the component CC of the capacity consumption EMD and the peak current component The Pearson correlation coefficient (PCC) is used to select the components whose correlation with the battery health status reference value is within a set range as battery health indicators (HIs). The set range is set according to actual conditions, such as -1 to -0.9 and 0.9 to 1.

[0079] Specifically, on the one hand, the capacity consumption of the target segment is subjected to empirical mode decomposition to obtain the components of the capacity consumption; on the other hand, the current peak of the target segment is subjected to empirical mode decomposition to obtain the components of the current peak. The components of the capacity consumption include: a plurality of intrinsic mode functions corresponding to the capacity consumption and a residual, such as Figure 3 As shown; the components of the current peak include: multiple intrinsic mode functions corresponding to the current peak and a residual, such as Figure 4 shown.

[0080] The EMD method combines CC and I max The data is continuously decomposed into multiple intrinsic mode functions (IMFs) and a residual error (RES) to capture its changing trend. The calculation formula is as follows:

[0081]

[0082] Where x(t) represents the original data (in this application, it refers to the capacity consumption CC and the current peak value I max ), c j (t) represents the jth IMF, r(t) represents RES, and J represents the number of IMFs.

[0083] The specific process of the EMD method is as follows:

[0084] 1. Extract extreme points: In EMD, first find the signal (i.e., capacity consumption CC and current peak I max ) may be the local maximum or minimum of the signal.

[0085] 2. Determine the upper and lower envelopes: Use interpolation to smooth the curves between the extreme points to obtain the upper and lower envelopes. These envelopes will help determine the IMFs.

[0086] 3. Extract IMF: Extract the first IMF from the signal, which is the average value between the local maximum and minimum in the signal.

[0087] 4. Repeat this process until all IMFs are extracted from the signal. Each time, subtract the IMF when extracting it until the remaining IMF is obtained.

[0088] 5. Subtract the original signal from all IMFs to obtain RES.

[0089] Afterwards, PCC was used to calculate the correlation coefficients of the original data, IMFs, RES, and the battery SOH reference value, and the components with greater correlation with the battery SOH reference value were selected to input into the model.

[0090] The PCC calculation formula is as follows:

[0091]

[0092] Among them, r xy represents the correlation coefficient, n represents the number of sampling points in the target segment, and x i represents the i-th value in the reference sequence, and in this application represents the marked battery SOH reference value, y i represents the ith value in the comparison sequence, which in this application are the original data, IMFs and RES, and represent the average values ​​of the reference sequence and the comparison sequence, respectively.

[0093] Step 6: Determine the remaining power sequence and the average temperature sequence according to the remaining power, maximum temperature and minimum temperature of the target segment, and use the Pearson correlation coefficient to select the sequence whose correlation with the battery health feature is within a set range as the feature influencing factor.

[0094] Specifically, on the one hand, the remaining power of the target segment, i.e., the state of capacity (SOC), is extracted to obtain the remaining power sequence; on the other hand, the maximum temperature and the minimum temperature of the target segment are extracted, and the formula is used and Calculate the average temperature T of the target fragment a , and get the average temperature series; where T i represents the average temperature of the i-th sampling point in the target segment, n represents the number of sampling points in the target segment, T imax represents the maximum temperature of the i-th sampling point in the target segment, T imin Indicates the minimum temperature of the i-th sampling point in the target segment.

[0095] Afterwards, PCC is used to calculate the correlation coefficients of the remaining power, average temperature, and battery health characteristics. The sequences with high correlation with the battery health characteristics are selected as input to the model.

[0096] Step 7: Construct an ensemble learning (EL) model, and train the ensemble learning model according to the battery health characteristics, characteristic influencing factors and corresponding battery health status reference values ​​of the target segment to obtain a trained ensemble learning model for evaluating battery SOH.

[0097] This step mainly includes: determining the number of basic models in the EL model, the offline training mode and the weight update strategy. In this application, the number of basic models is 4.

[0098] like Figure 5 As shown, the basic models for determining the EL model are two Gaussian process regression (GPR) models and two support vector machine (SVR) models. The offline training mode is: the first GPR is trained by the data of vehicle 1 collected by the big data platform, the second GPR is trained by the data of vehicle 2 collected by the big data platform, the first SVR is trained by the data of vehicle 1, and the second SVR is trained by the data of vehicle 2. The output of the integrated learning model is the weighted sum of the outputs of the four basic models. The weight update strategy is: every 5,000 kilometers, every 10,000 kilometers or every 20,000 kilometers, the weight of the basic model is updated according to the error of each basic model. The construction and weight update process of the EL model is as follows:

[0099] First, when a weight update event occurs, the error Error of each base model model The calculation is as follows:

[0100]

[0101] in, and model They represent the model output (i.e., the battery SOH prediction value) and the SOH true value (i.e., the battery SOH reference value) respectively. The battery SOH prediction value output by each basic model is different, and the input of each basic model is the battery health characteristics and the characteristic influencing factors related to the battery health characteristics. model represents the serial number of the basic model, β represents the number of basic models, and the value of β in this application is 4.

[0102] The importance coefficients of the four basic models are then calculated as follows:

[0103]

[0104] Among them, r model Represents the importance coefficient of the model-th basic model.

[0105] Then the obtained importance coefficient is normalized to get the weight ω of each basic model model :

[0106]

[0107] Finally, the output of the EL model, i.e., the SOH evaluation result h f It can be expressed as:

[0108]

[0109] Among them, h model (x f ) represents the output of the model-th basic model, x f Represents the input of the base model.

[0110] Step 8: Using the trained integrated learning model, the battery health state is evaluated based on the current driving data of the electric vehicle to obtain a battery health state prediction value.

[0111] Furthermore, the method also includes: drawing a battery degradation trajectory according to the battery health status prediction value.

[0112] In summary, the battery health status assessment method based on vehicle driving data provided by this application analyzes the driving data in the actual vehicle data, uses the EMD and PCC methods to assist in extracting the battery health factors (including battery health characteristics and characteristic influencing factors), and constructs an SOH assessment model based on the EL model to achieve accurate assessment of the battery SOH. Compared with the prior art, this application improves the utilization rate of the original data and avoids the consumption of manpower, material resources and financial resources caused by the need to design a large number of calibration experiments to evaluate the SOH of the actual vehicle power battery.

[0113] In an exemplary embodiment, the present application further provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0114] In an exemplary embodiment, the present application further provides a computer-readable storage medium storing a computer program, which implements the steps in the above-mentioned method embodiments when executed by a processor.

[0115] In an exemplary embodiment, the present application further provides a computer program product, including a computer program, which implements the steps in the above-mentioned method embodiments when executed by a processor.

[0116] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant legal provisions.

[0117] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0118] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0119] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0120] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A battery health status assessment method based on vehicle driving data, characterized in that: include: Obtain historical driving data of electric vehicles and retain corresponding reference data items; The reference data items include: vehicle speed, accumulated mileage, current, remaining power, latitude and longitude coordinates, maximum temperature, minimum temperature and vehicle identification code; Selecting a target segment from the historical driving data according to the vehicle identification code, latitude and longitude coordinates and vehicle driving speed in the reference data item; Calculating a battery health status reference value according to the cumulative mileage, maximum temperature and minimum temperature of the target segment; Calculating capacity consumption and current peak value according to the current of the target segment; Performing empirical mode decomposition on the capacity consumption and current peak of the target segment to obtain components of the capacity consumption and components of the current peak, and using the Pearson correlation coefficient to select components whose correlation with the battery health state reference value is within a set range as battery health features; Determine a remaining power sequence and an average temperature sequence according to the remaining power, maximum temperature, and minimum temperature of the target segment, and use the Pearson correlation coefficient to select a sequence whose correlation with the battery health feature is within a set range as a feature influencing factor; Constructing an integrated learning model, and training the integrated learning model according to the battery health characteristics, characteristic influencing factors and corresponding battery health status reference values ​​of the target segment to obtain a trained integrated learning model; The trained integrated learning model is used to evaluate the battery health state based on the current driving data of the electric vehicle to obtain a battery health state prediction value.

2. The battery health status assessment method based on vehicle driving data according to claim 1 is characterized in that: Obtain the historical driving data of electric vehicles and retain the corresponding reference data items, including: From the actual vehicle data of electric vehicles collected on the big data platform, the historical driving data of electric vehicles is screened based on the vehicle driving speed not being equal to 0, and the corresponding reference data items are retained.

3. The battery health status assessment method based on vehicle driving data according to claim 1, characterized in that: Selecting a target segment from the historical driving data according to the vehicle identification code, the latitude and longitude coordinates, and the vehicle driving speed in the reference data item includes: According to the vehicle identification code and the latitude and longitude coordinates, a data set having the same vehicle model and vehicle driving route is selected from the historical driving data; According to the longitude and latitude coordinates, a straight line track having a mileage span less than or equal to a set distance is selected from the data set for segmentation to obtain a plurality of discharge segments; The average vehicle speed of each of the discharge segments is calculated according to the vehicle running speed, the average vehicle speed interval is divided, and the discharge segment with the largest average vehicle speed interval proportion is selected as the target segment.

4. The battery health status assessment method based on vehicle driving data according to claim 1, characterized in that: Calculating a battery health status reference value according to the cumulative mileage, maximum temperature, and minimum temperature of the target segment, including: The battery health status calculation formula is SOH = 1.084M -0.01884 -0.04368+0.00572T m +1.657×10 -7 M m -0.0001524T m 2 +6.964×10 -9 T m M m , calculate the battery health state reference value SOH according to the cumulative mileage, maximum temperature and minimum temperature of the target segment; where M represents the cumulative mileage of the target segment, T m represents the monthly average temperature calculated based on the maximum and minimum temperatures of the target segment, M m Indicates the monthly cumulative mileage calculated based on the cumulative mileage of the target segment.

5. The battery health status assessment method based on vehicle driving data according to claim 1, characterized in that: Calculating the capacity consumption and the current peak value according to the current of the target segment includes: Using the capacity consumption calculation formula Calculate the capacity consumption CC according to the current of the target segment; wherein I(i) represents the current of the i-th sampling point in the training segment, and n represents the number of sampling points in the target segment; The current peak value is calculated using the formula I max =max(I(1), I(2), ..., I(n)), calculate the current peak value I according to the current of the target segment max ; Wherein, (I(1), I(2), ..., I(n)) represents the current of all sampling points in the target segment.

6. The battery health status assessment method based on vehicle driving data according to claim 1, characterized in that: Performing empirical mode decomposition on the capacity consumption and current peak of the target segment to obtain components of the capacity consumption and components of the current peak, including: Performing empirical mode decomposition on the capacity consumption of the target segment to obtain components of the capacity consumption; the components of the capacity consumption include: a plurality of intrinsic mode functions corresponding to the capacity consumption and a residual; Performing empirical mode decomposition on the current peak of the target segment to obtain components of the current peak; the components of the current peak include: a plurality of intrinsic mode functions corresponding to the current peak and a residual.

7. The battery health status assessment method based on vehicle driving data according to claim 1, characterized in that: Determining a remaining power sequence and an average temperature sequence according to the remaining power, the maximum temperature, and the minimum temperature of the target segment includes: Extracting the remaining power of the target segment to obtain a remaining power sequence; Extract the maximum and minimum temperatures of the target fragment and use the formula and Calculate the average temperature T of the target fragment a , and get the average temperature series; where T i represents the average temperature of the i-th sampling point in the target segment, n represents the number of sampling points in the target segment, T imax represents the maximum temperature of the i-th sampling point in the target segment, T imin Indicates the minimum temperature of the i-th sampling point in the target segment.

8. The battery health status assessment method based on vehicle driving data according to claim 1, characterized in that: The basic models of the integrated learning model include: two Gaussian process regression models and two support vector machine models; the output of the integrated learning model is the weighted sum of the outputs of the four basic models; during the training process, the weights of the four basic models are updated once every 5,000 kilometers, every 10,000 kilometers or every 20,000 kilometers according to the error of each basic model.

9. The battery health status assessment method based on vehicle driving data according to claim 1, characterized in that: Also includes: A battery degradation trajectory is plotted based on the battery health status prediction value.

10. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the battery health status assessment method based on vehicle driving data as described in any one of claims 1 to 9.

Images