A method and apparatus for predicting a state of health of a vehicle battery

Abstract

The application provides a vehicle battery health degree prediction method and device, wherein the method comprises the following steps: obtaining recorded data of a plurality of vehicle batteries under each charging condition, each piece of recorded data comprising battery data and driving data; obtaining target recorded data by filtering abnormal recorded data through quantile interval and quantile; calculating a single health degree corresponding to each target recorded data; performing linear fitting on the driving data and the single health degree in the target recorded data of each vehicle battery to obtain a linear fitting function corresponding to each vehicle battery; filtering vehicle batteries according to the linear fitting function corresponding to each vehicle battery to obtain target vehicle batteries; constructing a random forest model through the target recorded data in the target vehicle batteries and the single health degree corresponding to the target recorded data; and inputting recorded data of a vehicle battery to be predicted into the random forest model to predict the vehicle battery health degree of the vehicle battery to be predicted.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle battery maintenance, and particularly relates to a vehicle battery health degree prediction method and device. BACKGROUND

[0002] In the prior art, when the health degree of a vehicle battery is predicted, a large amount of battery data is acquired to construct a prediction model. Since the use of the vehicle battery is relatively complex, the battery data fluctuates greatly, and thus the prediction model constructed by the greatly fluctuating battery data has low accuracy in predicting the health degree of the vehicle battery. The low accuracy cannot help the user replace the vehicle battery in time, affecting the normal use of the vehicle, or causes the user to replace the vehicle battery too early, resulting in waste. SUMMARY

[0003] Therefore, the present application aims to at least provide a vehicle battery health degree prediction method and device, which first filters the charging data of a plurality of vehicle batteries, and then filters the vehicle batteries, to solve the technical problem of great data fluctuation in constructing a battery health degree prediction model in the prior art, and achieve the technical effect of improving the accuracy of predicting the battery health degree.

[0004] The present application mainly includes the following aspects:

[0005] In a first aspect, the present application provides a vehicle battery health degree prediction method, which includes: acquiring recorded data of a plurality of vehicle batteries under each charging condition, each piece of recorded data including battery data and driving data; filtering abnormal recorded data by quantile interval and quantile to obtain target recorded data; calculating a single health degree corresponding to each piece of target recorded data; performing linear fitting on the driving data and the single health degree in the target recorded data of each vehicle battery to obtain a linear fitting function corresponding to each vehicle battery; filtering the vehicle batteries according to the linear fitting function corresponding to each vehicle battery to obtain target vehicle batteries; constructing a random forest model by using the target recorded data in the target vehicle batteries and the single health degree corresponding to the target recorded data; and inputting the recorded data of a vehicle battery to be predicted into the random forest model to predict the vehicle battery health degree of the vehicle battery to be predicted.

[0006] Optionally, the target record data is obtained by quantile interval and quantile filtering of the abnormal record data, including: arranging the battery data and the driving data in ascending order respectively, determining a first quantile and a first quantile interval corresponding to the battery data, and determining a second quantile and a second quantile interval corresponding to the driving data; calculating a battery data interval corresponding to the first quantile and the first quantile interval; calculating a driving data interval corresponding to the second quantile and the second quantile interval; determining record data in which the battery data belongs to the battery data interval and the driving data belongs to the driving data interval as the target record data.

[0007] Optionally, the battery data includes: charging duration, charging current, initial state of charge and final state of charge, and the single health degree corresponding to each target record data is calculated, including: calculating the integral of the charging current corresponding to each target record data in the charging duration, and determining the result after the integral as the charging capacity corresponding to each target record data; determining the single health degree corresponding to each target record data according to the charging capacity, the initial state of charge and the final state of charge.

[0008] Optionally, the single health degree corresponding to each target record data is determined according to the charging capacity, the initial state of charge and the final state of charge, including: obtaining the calibrated capacity of the vehicle battery; calculating the absolute value of the difference between the final state of charge and the initial state of charge, and comparing the charging capacity with the absolute value, and determining the ratio as the total capacity of the battery corresponding to each target record data; comparing the calibrated capacity with the total capacity of the battery, and determining the ratio as the single health degree corresponding to each target record data.

[0009] Optionally, the driving data and the single health degree in the target record data of each vehicle battery are linearly fitted to obtain a linear fitting function corresponding to each vehicle battery, including: linearly fitting the driving data in the target record data of each vehicle battery as the independent variable and the single health degree corresponding to the target record data of each vehicle battery as the dependent variable to obtain a linear fitting function corresponding to each vehicle battery.

[0010] Optionally, the vehicle battery is filtered according to the linear fitting function corresponding to each vehicle battery to obtain a target vehicle battery, including: obtaining the confidence, slope and intercept of each linear fitting function; filtering the abnormal linear fitting function by quantile interval and quantile to obtain a target linear fitting function; and determining the vehicle battery corresponding to the target linear fitting function as the target vehicle battery.

[0011] Optionally, the target linear fitting function is obtained by linear fitting functions of quantile interval and quantile filtering outliers, comprising: arranging the confidence, slope and intercept of each linear fitting function in ascending order respectively, determining the characteristic third quantile and the characteristic third quantile interval corresponding to the confidence, determining the characteristic fourth quantile and the characteristic fourth quantile interval corresponding to the slope, and determining the characteristic fifth quantile and the characteristic fifth quantile interval corresponding to the intercept; calculating the confidence interval corresponding to the characteristic third quantile and the characteristic third quantile interval; calculating the slope interval corresponding to the characteristic fourth quantile and the characteristic fourth quantile interval; calculating the intercept interval corresponding to the characteristic fifth quantile and the characteristic fifth quantile interval; determining the linear fitting function whose confidence belongs to the confidence interval, whose slope belongs to the slope interval and whose intercept belongs to the intercept interval as the target linear fitting function.

[0012] In a second aspect, the embodiments of the present application further provide a device for predicting the state of health of a vehicle battery. The device comprises: an acquisition module configured to acquire recorded data of a plurality of vehicle batteries under each charging condition, each piece of recorded data comprising battery data and driving data; a first determination module configured to obtain target recorded data by filtering abnormal recorded data based on quantile interval and quantile; a calculation module configured to calculate a single-time state of health corresponding to each piece of target recorded data; a second determination module configured to perform linear fitting on the driving data and the single-time state of health in the target recorded data of each vehicle battery to obtain a linear fitting function corresponding to each vehicle battery; a third determination module configured to filter vehicle batteries based on the linear fitting function corresponding to each vehicle battery to obtain target vehicle batteries; a construction module configured to construct a random forest model based on the target recorded data in the target vehicle batteries and the single-time state of health corresponding to the target recorded data; and a prediction module configured to input recorded data of a vehicle battery to be predicted into the random forest model to predict the state of health of the vehicle battery to be predicted.

[0013] In a third aspect, the embodiments of the present application further provide an electronic device, comprising: a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the electronic device is running, the processor and the memory communicate through the bus, and the machine readable instructions are executed by the processor to perform the steps of the prediction method of the state of health of a vehicle battery in the first aspect or any possible implementation manner of the first aspect.

[0014] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by the processor to perform the steps of the prediction method of the state of health of a vehicle battery in the first aspect or any possible implementation manner of the first aspect.

[0015] The embodiment of the present application provides a kind of vehicle battery health degree prediction method and device, obtain the record data of multiple vehicle batteries in each charging condition, each record data includes battery data, travel data;By quantile interval and quantile filter abnormal record data, obtain target record data;The single health degree corresponding to each target record data is calculated;Linear fitting is carried out to the travel data in the target record data of each vehicle battery and single health degree, and the linear fitting function corresponding to each vehicle battery is obtained;According to the linear fitting function corresponding to each vehicle battery, target vehicle battery is obtained by filtering vehicle battery;Random forest model is constructed by the target record data in target vehicle battery and the single health degree corresponding to target record data;The record data of the vehicle battery to be predicted is input into random forest model, and the vehicle battery health degree of the vehicle battery to be predicted is predicted.The present application is filtered to the charging data of multiple vehicle batteries first, then the vehicle battery is filtered, the technical problem that data fluctuation is larger in the prior art for constructing battery health degree prediction model is solved, and the technical effect that the accuracy of predicting battery health degree is improved is achieved.

[0016] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the following preferred embodiments are described in detail below, and the accompanying drawings are described as follows. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0018] Figure 1 A flow chart of a vehicle battery health degree prediction method provided by the embodiment of the present application is shown.

[0019] Figure 2 A flow chart of the step of obtaining target record data by filtering abnormal record data through quantile interval and quantile provided by the embodiment of the present application is shown.

[0020] Figure 3 A functional module diagram of a vehicle battery health degree prediction device provided by the embodiment of the present application is shown.

[0021] Figure 4 A structure diagram of an electronic device provided by the embodiment of the present application is shown. DETAILED DESCRIPTION

[0022] In order to make the purposes, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of description and illustration, and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn according to the actual proportions. The flowcharts used in the present application show the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowcharts can not be implemented in sequence, and the steps without logical context relationship can be reversed in sequence or implemented simultaneously. In addition, one or more other operations can be added to the flowcharts or one or more operations can be removed from the flowcharts under the guidance of the content of the present application.

[0023] In addition, the described embodiments are only some of the embodiments of the present application, not all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0024] In the prior art, all charging data of the battery during the charging process is used to build a prediction model of the vehicle battery health, which causes some abnormal data to also participate in modeling, so that the accuracy of the predicted vehicle battery health is not high, thereby affecting the user's judgment of whether the vehicle battery should be replaced.

[0025] Based on this, the embodiments of the present application provide a vehicle battery health prediction method and device, which first filters the charging data of multiple vehicle batteries, and then filters the vehicle batteries, solving the technical problem of large data fluctuation in building a battery health prediction model in the prior art, achieving the technical effect of improving the accuracy of predicting battery health, as follows:

[0026] Please refer to Figure 1 , Figure 1 The flowchart of a vehicle battery health prediction method provided by the embodiments of the present application is shown in FIG. 1. As shown in FIG. 1, the vehicle battery health prediction method provided by the embodiments of the present application includes the following steps: Figure 1

[0027] S101, record data of multiple vehicle batteries under each charging condition is obtained, and each record data includes battery data and driving data.

[0028] ​The record data of each charging of the plurality of vehicle batteries is obtained, that is, the record data of the vehicle battery corresponding to the vehicle in the non-driving state and the vehicle battery in the charging state is obtained. The plurality of vehicle batteries are vehicle batteries of the same series of vehicles, that is, the random forest model is used to predict the battery health of the vehicle batteries of the series of vehicles.

[0029] Each record data includes battery data, driving data, and record identification. The battery data is used to describe the charging condition of the vehicle battery, the driving data is used to describe the driving condition of the vehicle corresponding to the vehicle battery, and the record identification is used to describe the battery number and the charging number corresponding to the record data. The battery data includes charging duration, charging current, initial state of charge, final state of charge, cumulative charging amount, cumulative discharging amount, and fast charging ratio.

[0030] The charging duration refers to the charging duration corresponding to each record identification; the charging current refers to the charging current in the charging duration range corresponding to each record identification; the initial state of charge refers to the state of charge of the vehicle battery at the initial moment of the charging duration corresponding to each record identification; the final state of charge refers to the state of charge of the vehicle battery at the end moment of the charging duration corresponding to each record identification; the cumulative charging amount refers to the cumulative charging amount of each vehicle battery up to the record identification; the cumulative discharging amount refers to the cumulative discharging amount of each vehicle battery up to the record identification; and the fast charging ratio refers to the ratio of the total fast charging number of the vehicle battery up to the record identification to the charging number of the vehicle battery corresponding to the record identification.

[0031] The charging current and the charging duration are obtained by using the battery swap advantage to obtain the high-frequency and high-precision charging current and charging duration of the vehicle battery in the battery swap station. The cumulative charging amount, the cumulative discharging amount, and the fast charging ratio are determined by data profiling from the data reported by GBT32960.

[0032] For example, the record identification corresponding to the third charging of the first block of vehicle batteries can be set as 001003, that is, the first three digits represent the battery number, and the last three digits represent the charging number of the battery. The record number 001003 corresponds to the third charging duration, the third charging current, the initial state of charge of the vehicle battery at the third charging, the final state of charge of the vehicle battery at the third charging, the cumulative charging amount of the first three chargings, the cumulative discharging amount of the first three chargings, and the fast charging ratio of the first three chargings.

[0033] In S102, the target record data is obtained by filtering the record data by quantile interval and quantile.

[0034] The battery data is arranged in ascending order, that is, the battery data in all record data is arranged in ascending order.

[0035] Referring to Figure 2 , Figure 2 The flow chart of the step of obtaining target record data by filtering abnormal record data through quantile interval and quantile provided by the embodiment of the application. As Figure 2 shown, the step of obtaining target record data by filtering abnormal record data through quantile interval and quantile provided by the embodiment of the application includes the following steps:

[0036] S1021, arrange the battery data and the driving data in ascending order respectively, determine the feature one quantile and the feature one quantile interval corresponding to the battery data, and determine the feature two quantile and the feature two quantile interval corresponding to the driving data.

[0037] The quantile adopted by the embodiment of the application is quartile, and the quantile interval is quartile interval. The selection of the quantile interval and the quantile is not limited here, and can be adjusted according to actual conditions.

[0038] Specifically, the 25th battery data, the 50th battery data and the 75th battery data after ascending arrangement are determined as the feature one quantile corresponding to the battery data. That is, the 25th battery data after ascending arrangement is determined as the lower quartile in the feature one quantile corresponding to the battery data; the 50th battery data after ascending arrangement is determined as the median in the feature one quantile corresponding to the battery data; and the 75th battery data after ascending arrangement is determined as the upper quartile in the feature one quantile corresponding to the battery data. The difference between the 75th battery data and the 25th battery data is determined as the feature one quantile interval corresponding to the battery data, that is, the difference between the upper quartile in the feature one quantile and the lower quartile in the feature one quantile is determined as the feature one quantile interval corresponding to the battery data.

[0039] For example, if there are 100 battery data, and each battery data includes: charging duration, charging current, initial state of charge, end state of charge, cumulative charge amount, cumulative discharge amount and fast charging ratio. Then arrange the 100 charging durations in ascending order, and then determine the 25th charging duration, the 50th charging duration and the 75th charging duration after ascending arrangement as the first quartile corresponding to the charging duration, and determine the difference between the 75th charging duration and the 25th charging duration after ascending arrangement as the feature one quantile interval corresponding to the charging duration. The charging current, the initial state of charge, the end state of charge, the cumulative charge amount, the cumulative discharge amount and the fast charging ratio are also operated as described above to determine the respective feature one quantile and feature one quantile interval.

[0040] S1022, calculate the battery data interval corresponding to the feature one quantile and the feature one quantile interval; calculate the driving data interval corresponding to the feature two quantile and the feature two quantile interval.

[0041] The left interval endpoint of the battery data interval is calculated by the following formula:

[0042] lowlimit1=a1-1.5×d1 (1)

[0043] In formula (1), lowlimit1 refers to the left interval endpoint of the battery data interval, a1 refers to the lower quartile of the feature one quantile of the battery data, and d1 refers to the feature one quantile interval of the battery data. The lower quartile of the feature one quantile of the battery data is subtracted by 1.5 times the feature one quantile interval, and the difference is determined as the left interval endpoint of the battery data interval.

[0044] The right interval endpoint of the battery data interval is calculated by the following formula:

[0045] uplimit1=c1+1.5×d1 (2)

[0046] In formula (2), uplimit1 refers to the right interval endpoint of the battery data interval, c1 refers to the upper quartile of the feature one quantile of the battery data, and d1 refers to the feature one quantile interval of the battery data. The upper quartile of the feature one quantile of the battery data is added by 1.5 times the feature one quantile interval, and the sum is determined as the right interval endpoint of the battery data interval.

[0047] The left interval endpoint of the driving data interval is calculated by the following formula:

[0048] lowlimit2=a2-1.5×d2 (3)

[0049] In formula (3), lowlimit2 refers to the left interval endpoint of the driving data interval, a2 refers to the lower quartile of the feature two quantile of the driving data, and d2 refers to the feature two quantile interval of the driving data. The lower quartile of the feature two quantile of the driving data is subtracted by 1.5 times the feature two quantile interval, and the difference is determined as the left interval endpoint of the driving data interval.

[0050] The right interval endpoint of the driving data interval is calculated by the following formula:

[0051] uplimit2=c2+1.5×d2 (4)

[0052] In formula (4), uplimit2 refers to the right interval end point of the travel data interval, c2 refers to the upper quartile of the feature biquartiles of the travel data, d2 refers to the feature biquartile interval of the travel data, the upper quartile of the feature biquartiles of the travel data is summed with 1.5 times the feature biquartile interval, and the sum value is determined as the right interval end point of the travel data interval.

[0053] S1023, the record data in which the battery data belongs to the battery data interval and the travel data belongs to the travel data interval is determined as the target record data.

[0054] It is judged whether the battery data belongs to the battery data interval. If the battery data belongs to the battery data interval, the record data in which the battery data belongs to the battery data interval is determined as the intermediate record data. If the battery data does not belong to the battery data interval, the corresponding record data is deleted.

[0055] It is judged whether the travel data of the intermediate record data belongs to the travel data interval. If the travel data of the intermediate record data belongs to the travel data interval, the intermediate record data in which the travel data belongs to the travel data interval is determined as the target record data. If the travel data of the intermediate record data does not belong to the travel data interval, the corresponding intermediate record data is deleted.

[0056] That is, the battery data and the travel data of each target record data belong to the corresponding battery data interval and travel data interval respectively.

[0057] Return Figure 1 , S103, the single health degree corresponding to each target record data is calculated.

[0058] The single health degree is used to describe the health degree of each vehicle battery under this charging condition.

[0059] The single health degree corresponding to each target record data is calculated, including:

[0060] The integral of the charging current corresponding to each target record data in the charging duration is calculated, and the result after integration is determined as the charging capacity corresponding to each target record data; according to the charging capacity, the initial state of charge and the final state of charge, the single health degree corresponding to each target record data is determined.

[0061] The charging capacity corresponding to each target record data is determined by the following formula:

[0062]

[0063] In formula (5), capacity iThe target record data corresponds to the charging capacity, i refers to the record identification corresponding to each target record data, t refers to the charging time corresponding to each target record data, and I refers to the charging current corresponding to each target record data. The integral of the charging current corresponding to each target record data in the charging time is calculated, and the result after integration is determined as the charging capacity corresponding to each target record data.

[0064] According to the charging capacity, the initial state of charge and the end state of charge, the single health degree corresponding to each target record data is determined, including:

[0065] The calibrated capacity of the vehicle battery is obtained; the absolute value of the difference between the end state of charge and the initial state of charge is calculated, the charging capacity is compared with the absolute value, and the ratio is determined as the total capacity of the battery corresponding to each target record data; the calibrated capacity is compared with the total capacity of the battery, and the ratio is determined as the single health degree corresponding to each target record data.

[0066] The calibrated capacity refers to the calibrated capacity of the vehicle battery when it is manufactured.

[0067] The total capacity of the battery corresponding to each target record data is determined by the following formula:

[0068]

[0069] In formula (6), totalcapacity i The single health degree corresponding to each target record data is soh, endsoc i The end state of charge corresponding to each target record data is Startsoc i The initial state of charge corresponding to each target record data is calculated, the absolute value of the difference between the end state of charge and the initial state of charge is calculated, the charging capacity is compared with the absolute value, and the ratio is determined as the total capacity of the battery corresponding to each target record data.

[0070] The single health degree corresponding to each target record data is determined by the following formula:

[0071]

[0072] In formula (7), soh i The single health degree corresponding to each target record data is soh, capacity0 refers to the calibrated capacity of the vehicle battery. The calibrated capacity is compared with the total capacity of the battery, and the ratio is determined as the single health degree corresponding to each target record data.

[0073] S104, the driving data and the single health degree in the target record data of each vehicle battery are linearly fitted, and the linear fitting function corresponding to each vehicle battery is obtained.

[0074] The driving data in the target record data of each vehicle battery is linearly fitted with the single health degree, to obtain a linear fitting function corresponding to each vehicle battery, including:

[0075] The driving data in the target record data of each vehicle battery is linearly fitted as the independent variable, and the single health degree corresponding to the target record data of each vehicle battery is linearly fitted as the dependent variable, to obtain a linear fitting function corresponding to each vehicle battery.

[0076] Specifically, the driving data includes driving mileage and purchase date. The driving mileage refers to the total driving mileage data of the vehicle corresponding to the driving data until each charging condition, and the purchase date refers to the total use date of the vehicle corresponding to the driving data until each charging condition. In the embodiment of the application, the starting date of the use date of the vehicle is the date when the user purchases the vehicle.

[0077] That is, the driving mileage in the target record data of each vehicle battery is taken as the first independent variable, the purchase date in the target record data of each vehicle battery is taken as the second independent variable, and the single health degree corresponding to the target record data of each vehicle battery is taken as the dependent variable, to obtain a linear fitting function corresponding to each vehicle battery.

[0078] That is, a plurality of target record data and single health degrees corresponding to one vehicle battery, and one linear fitting function corresponding to one vehicle battery.

[0079] S105, filtering the vehicle battery according to the linear fitting function corresponding to each vehicle battery to obtain a target vehicle battery.

[0080] According to the linear fitting function corresponding to each vehicle battery, the vehicle battery is filtered to obtain a target vehicle battery, including:

[0081] Obtaining the confidence, slope and intercept of each linear fitting function; filtering the abnormal linear fitting function through quantile interval and quantile to obtain a target linear fitting function.

[0082] Filtering the abnormal linear fitting function through quantile interval and quantile to obtain a target linear fitting function, including:

[0083] The confidence, slope and intercept of each linear fitting function are arranged in ascending order respectively, the feature tertiles corresponding to the confidence and the feature tertile interval are determined, the feature quartiles corresponding to the slope and the feature quartile interval are determined, and the feature quintiles corresponding to the intercept and the feature quintile interval are determined; the confidence interval corresponding to the feature tertiles and the feature tertile interval is calculated; the slope interval corresponding to the feature quartiles and the feature quartile interval is calculated; the intercept interval corresponding to the feature quintiles and the feature quintile interval is calculated; the linear fitting function whose confidence belongs to the confidence interval, whose slope belongs to the slope interval and whose intercept belongs to the intercept interval is determined as the target linear fitting function; and the vehicle battery corresponding to the target linear fitting function is determined as the target vehicle battery.

[0084] Specifically, the ways of determining the feature tertiles, the feature tertile interval and the confidence interval, the feature quartiles, the feature quartile interval and the slope interval, and the feature quintiles, the feature quintile interval and the intercept interval are the same as the method of determining the feature first quantile, the feature first quantile interval and the battery data interval, which will not be described here.

[0085] The linear fitting function whose confidence belongs to the confidence interval, whose slope belongs to the slope interval and whose intercept belongs to the intercept interval is determined as the target linear fitting function; and the vehicle battery corresponding to the target linear fitting function is determined as the target vehicle battery.

[0086] That is, the linear fitting function whose confidence does not belong to the confidence interval, or whose slope does not belong to the slope interval, or whose intercept does not belong to the intercept interval is determined as the target linear fitting function.

[0087] S106, constructing a random forest model through the target record data in the target vehicle battery and the single-time health degree corresponding to the target record data.

[0088] Constructing a random forest model through the driving data in the target record data in the target vehicle battery and the single-time health degree corresponding to the target record data.

[0089] The reason for using the driving data is that when the user uses the random forest model for prediction, the user can only determine the driving data (driving mileage, purchase date), and cannot know the battery data. Therefore, when constructing the random forest model, only the driving data in the target record data in the target vehicle battery and the single-time health degree corresponding to the target record data are used.

[0090] S107, inputting the record data of the vehicle battery to be predicted into the random forest model to predict the vehicle battery health degree of the vehicle battery to be predicted.

[0091] The user inputs the record data of the vehicle battery to be predicted into the random forest model when predicting the vehicle battery health, and predicts the vehicle battery health of the vehicle battery to be predicted.

[0092] Based on the same application concept, the embodiment of the present application also provides a vehicle battery health prediction device corresponding to the vehicle battery health prediction method provided by the above-embodiment. Since the principle of solving problems in the device of the embodiment of the present application is similar to the vehicle battery health prediction method of the above-embodiment of the present application, the implementation of the device can be referred to the implementation of the method, and the repeated parts will not be described here.

[0093] As shown in Figure 3 , Figure 3 A functional module diagram of a vehicle battery health prediction device provided by the embodiment of the present application. A vehicle battery health prediction device 10 comprises an acquisition module 101, a first determination module 102, a calculation module 103, a second determination module 104, a third determination module 105, a construction module 106 and a prediction module 107. The acquisition module 101 is configured to acquire record data of a plurality of vehicle batteries under each charging condition, and each record data comprises battery data and driving data. The first determination module 102 is configured to filter out target record data from the record data by quantile interval and quantile filtering. The calculation module 103 is configured to calculate a single health corresponding to each target record data. The second determination module 104 is configured to perform linear fitting on the driving data and the single health in the target record data of each vehicle battery, and obtain a linear fitting function corresponding to each vehicle battery. The third determination module 105 is configured to filter the vehicle battery according to the linear fitting function corresponding to each vehicle battery to obtain a target vehicle battery. The construction module 106 is configured to construct a random forest model by using the target record data in the target vehicle battery and the single health corresponding to the target record data. The prediction module 107 is configured to input the record data of a vehicle battery to be predicted into the random forest model, and predict the vehicle battery health of the vehicle battery to be predicted.

[0094] Based on the same application concept, referring to Figure 4 Fig. 2 shows a structural schematic diagram of an electronic device provided by the embodiment of the present application. The electronic device 20 comprises a processor 201, a memory 202 and a bus 203. The memory 202 stores machine readable instructions executable by the processor 201. When the electronic device 20 is running, the processor 201 and the memory 202 communicate through the bus 203. When the machine readable instructions are executed by the processor 201, the steps of the vehicle battery health prediction method of any one of the above-embodiments are performed.

[0095] Specifically, the machine readable instructions, when executed by the processor 201, can perform the following processing: obtaining record data of a plurality of vehicle batteries in each charging case, each record data including battery data and driving data; obtaining target record data by filtering out abnormal record data through quantile interval and quantile; calculating a single health degree corresponding to each target record data; performing linear fitting on the driving data and the single health degree in the target record data of each vehicle battery to obtain a linear fitting function corresponding to each vehicle battery; filtering vehicle batteries according to the linear fitting function corresponding to each vehicle battery to obtain target vehicle batteries; constructing a random forest model through the target record data in the target vehicle batteries and the single health degree corresponding to the target record data; and inputting record data of a vehicle battery to be predicted into the random forest model to predict a vehicle battery health degree of the vehicle battery to be predicted.

[0096] Based on the same application concept, the embodiments of the present application also provide a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by a processor to perform the steps of the vehicle battery health degree prediction method provided in the above embodiments.

[0097] Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, etc., and the computer program stored on the storage medium can be executed to perform the vehicle battery health degree prediction method. By first screening the charging data of a plurality of vehicle batteries and then screening the vehicle batteries, the technical problem of large data fluctuation in constructing a battery health degree prediction model in the prior art is solved, and the technical effect of improving the accuracy of predicting the battery health degree is achieved.

[0098] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system and device can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here. In the several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented by other ways. The above-described device embodiments are only schematic, for example, the division of the units is only a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the shown or discussed mutual elements can be indirect coupling or communication connection through some communication interfaces, devices or units, and can be electrical, mechanical or other forms.

[0099] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0100] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present alone, or two or more units can be integrated in one unit.

[0101] If the functions are realized in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer readable storage medium executable by a processor. Based on this understanding, the technical solutions of the present application or the parts of the present application that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of software products. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, and various program code storage media.

[0102] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for predicting the health of a vehicle battery, characterized in that, The prediction method includes: Acquire recorded data from multiple vehicle batteries during each charge cycle. Each recorded data entry includes battery data and driving data. The target record data is obtained by filtering out abnormal record data using quantile intervals and quantiles; Calculate the single health score corresponding to each target record data, wherein the single health score indicates the health score of each vehicle battery under one charge condition; For each vehicle battery, the driving data in the target recorded data is linearly fitted with the single health status to obtain the linear fitting function corresponding to each vehicle battery. The target vehicle battery is obtained by filtering the vehicle batteries according to the linear fitting function corresponding to each vehicle battery. A random forest model is constructed using the driving data of the target recorded data in the target vehicle's battery and the single health score corresponding to the target recorded data. The driving data recorded by the vehicle battery to be predicted is input into the random forest model to predict the battery health of the vehicle. The step of obtaining target record data by filtering out abnormal record data through quantile intervals and quantiles includes: The battery data and the driving data are sorted in ascending order to determine the first quantile and the interval between the first quantiles of the battery data, and the second quantile and the interval between the second quantiles of the driving data. Calculate the battery data intervals corresponding to the first quantile of the feature and the interval between the first quantiles of the feature; Calculate the driving data intervals corresponding to the characteristic quartiles and the intervals between the characteristic quartiles; Recorded data in which both the battery data and the driving data belong to the battery data range are identified as target recorded data.

2. The prediction method according to claim 1, characterized in that, The battery data includes: charging duration, charging current, initial state of charge, and final state of charge. The calculation of the single-time health score corresponding to each target recorded data point includes: Calculate the integral of the charging current corresponding to each target record data over the charging time, and determine the charging capacity corresponding to each target record data based on the integral result. Based on the charging capacity, the initial state of charge, and the final state of charge, determine the single health level corresponding to each target recorded data.

3. The prediction method according to claim 2, characterized in that, The step of determining the single-time health score corresponding to each target recorded data based on the charging capacity, the initial state of charge, and the final state of charge includes: Obtain the calibrated capacity of the vehicle battery; Calculate the absolute value of the difference between the end state of charge and the beginning state of charge, compare the charging capacity with the absolute value, and determine the total battery capacity corresponding to each target recorded data by the ratio. The calibrated capacity is compared with the total battery capacity, and the ratio is used to determine the single health score corresponding to each target recorded data.

4. The prediction method according to claim 1, characterized in that, The method of linearly fitting the driving data in the target recorded data for each vehicle battery with the single health status to obtain the linear fitting function corresponding to each vehicle battery includes: Using the driving data in the target record data of each vehicle battery as the independent variable and the single health status corresponding to the target record data of each vehicle battery as the dependent variable, a linear fitting is performed to obtain the linear fitting function corresponding to each vehicle battery.

5. The prediction method according to claim 1, characterized in that, The step of filtering vehicle batteries to obtain the target vehicle battery based on the linear fitting function corresponding to each vehicle battery includes: Obtain the confidence level, slope, and intercept for each of the linear fitting functions; The target linear fitting function is obtained by filtering out anomalies using quantile intervals and quantiles. The vehicle battery corresponding to the target linear fitting function is determined as the target vehicle battery.

6. The prediction method according to claim 5, characterized in that, The linear fitting function obtained by filtering out anomalies through quantile intervals and quantiles to obtain the target linear fitting function includes: The confidence level, slope, and intercept of each linear fitting function are sorted in ascending order to determine the characteristic ternary and characteristic ternary intervals corresponding to the confidence level, the characteristic quartile and characteristic quartile intervals corresponding to the slope, and the characteristic quintile and characteristic quintile intervals corresponding to the intercept. Calculate the confidence intervals corresponding to the feature tertiles and the intervals between the feature tertiles; Calculate the slope intervals corresponding to the characteristic quartiles and the interquartile ranges; Calculate the intercept intervals corresponding to the characteristic quintiles and the intervals between the characteristic quintiles; The linear fitting function whose confidence level belongs to the confidence level interval, whose slope belongs to the slope interval, and whose intercept belongs to the intercept interval is determined as the target linear fitting function.

7. A device for predicting the health of a vehicle battery, characterized in that, The prediction device includes: The acquisition module is used to acquire recorded data from multiple vehicle batteries during each charge. Each record includes battery data and driving data. The first determining module is used to filter out abnormal record data by quantile interval and quantile to obtain target record data; The calculation module is used to calculate the single health score corresponding to each target recorded data, wherein the single health score indicates the health score of each vehicle battery under one charge condition; The second determining module is used to perform linear fitting between the driving data in the target recorded data of each vehicle battery and the single health status to obtain the linear fitting function corresponding to each vehicle battery. The third determination module is used to filter vehicle batteries according to the linear fitting function corresponding to each vehicle battery to obtain the target vehicle battery. The construction module is used to build a random forest model using the driving data of the target recorded data in the target vehicle's battery and the single health score corresponding to the target recorded data. The prediction module is used to input the driving data of the recorded data of the vehicle battery to be predicted into the random forest model to predict the vehicle battery health. The first determining module is further configured to sort the battery data and the driving data in ascending order, determine the first quantile and the interval between the first quantiles of the battery data, and determine the second quantile and the interval between the second quantiles of the driving data; calculate the battery data interval corresponding to the first quantile and the interval between the first quantiles of the battery data; calculate the driving data interval corresponding to the second quantile and the interval between the second quantiles of the battery data; and determine the recorded data in which the battery data belongs to the battery data interval and the driving data belongs to the driving data interval as target recorded data.

8. An electronic device, characterized in that, include: The device includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus. The machine-readable instructions are executed by the processor to perform the steps of the vehicle battery health prediction method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the vehicle battery health prediction method as described in any one of claims 1 to 6.

Images