CN115327382A

CN115327382A - Method and device for generating battery service life prediction model and vehicle

Info

Publication number: CN115327382A
Application number: CN202210784293.1A
Authority: CN
Inventors: 杨蕊; 吴怀仁; 黄忠山
Original assignee: Guangzhou Automobile Group Co Ltd
Current assignee: Guangzhou Automobile Group Co Ltd
Priority date: 2022-06-28
Filing date: 2022-06-28
Publication date: 2022-11-11
Anticipated expiration: 2042-06-28
Also published as: CN115327382B

Abstract

The embodiment of the application provides a method and a device for generating a battery service life prediction model and a vehicle, and relates to the technical field of battery prediction. The method comprises the steps of acquiring real data and experimental sample data of preset types; determining a functional relation between target sample data and the residual service life of the battery according to the experimental sample data; determining target data from the real data, wherein the target data and the target sample data belong to the same type of data; according to the real data, the target sample data and the functional relation, a service life prediction model for predicting the remaining service life of the battery is generated, so that the generated service life prediction model has rich data base and strong relevance with the real data, and the accuracy of the service life prediction model for predicting the remaining service life of the battery is improved.

Description

Method and device for generating battery service life prediction model and vehicle

Technical Field

The embodiment of the application relates to the technical field of battery prediction, in particular to a method and a device for generating a battery service life prediction model and a vehicle.

Background

At present, the electric quantity of a battery of a key of equipment is generally pre-warned by adopting a voltage threshold comparison pre-warning mode, namely, when the voltage is lower than the voltage pre-warning threshold, a low-electric-quantity prompt is triggered to remind a user to replace the battery. The voltage threshold comparison method adopts single voltage early warning, and cannot accurately predict the electric quantity of the key.

Disclosure of Invention

The embodiment of the application provides a method and a device for generating a battery service life prediction model and a vehicle, so as to solve the problems.

In a first aspect, an embodiment of the present application provides a method for generating a battery service life prediction model. The method comprises the following steps: acquiring real data and experimental sample data of a preset type; determining a functional relation between target sample data and the residual service life of the battery according to the experimental sample data; determining target data from the real data, wherein the target data and the target sample data belong to the same type of data; and generating a service life prediction model according to the real data, the target sample data and the functional relation, wherein the service life prediction model is used for predicting the residual service life of the battery.

In a second aspect, an embodiment of the present application provides a method for predicting a remaining service life of a battery. The method comprises the following steps: acquiring current real data of a preset type; and predicting the residual service life of the current battery based on the service life prediction model established according to the method provided by the first aspect of the embodiment of the application.

In a third aspect, an embodiment of the present application provides an apparatus for generating a battery service life prediction model. The device comprises: the data acquisition module is used for acquiring real data and experimental sample data of preset types, wherein the real data is data acquired by equipment; the function determining module is used for determining the functional relation between target sample data and the residual service life of the battery according to the experimental sample data; the data extraction module is used for determining target data from the real data, and the target data and the target sample data belong to the same type of data; and the model generation module is used for generating a service life prediction model according to the real data, the target sample data and the functional relation, and the service life prediction model is used for predicting the residual service life of the battery.

In a fourth aspect, an embodiment of the present application provides an apparatus for predicting remaining service life of a battery. The device includes: the data acquisition module is used for acquiring current real data of a preset type; the service life prediction module is used for predicting the remaining service life of the current battery based on the service life prediction model established according to the method provided by the first aspect of the embodiment of the application.

In a fifth aspect, embodiments of the present application provide a vehicle. The vehicle includes memory, one or more processors, and one or more applications. Wherein one or more application programs are stored in the memory and configured to perform the methods provided by the embodiments of the present application when invoked by the one or more processors.

The embodiment of the application provides a method and a device for generating a battery service life prediction model and a vehicle, the service life prediction model is generated by adopting preset type real data and experiment sample data, the relation between the experiment data and the real data is considered while the residual service life of the battery is predicted through a large amount of abundant data, so that the generated service life prediction model has abundant data bases and strong relevance with the real data, and the accuracy of the residual service life prediction of the battery by the service life prediction model is improved.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a schematic diagram of an application scenario of a method provided in an embodiment of the present application, provided in an exemplary embodiment of the present application;

fig. 2 is a schematic flowchart of a method for generating a battery service life prediction model according to an embodiment of the present disclosure;

FIG. 3 is a method for generating a battery life prediction model according to another embodiment of the present application;

fig. 4 is a schematic flowchart of a method for predicting remaining service life of a battery according to an embodiment of the present disclosure;

FIG. 5 is a schematic flowchart illustrating a method for predicting remaining battery life according to another embodiment of the present disclosure;

FIG. 6 is a flowchart illustrating a method for predicting remaining useful life of a battery according to an exemplary embodiment of the present disclosure;

fig. 7 is a block diagram of a device for generating a battery service life prediction model according to an embodiment of the present application;

fig. 8 is a block diagram of a device for predicting remaining service life of a battery according to an embodiment of the present disclosure;

FIG. 9 is a block diagram of a vehicle according to an embodiment of the present application;

fig. 10 is a block diagram of a computer-readable storage medium according to an embodiment of the present disclosure.

Detailed Description

In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application.

Referring to fig. 1, fig. 1 is a schematic view of an application scenario of a method provided in an embodiment of the present application according to an exemplary embodiment of the present application. The service life prediction system 10 includes a life prediction module 11, a model generation module 12, and a data storage module 13. The life prediction module 11, the model generation module 12 and the data storage module 13 have communication links with each other, so that data interaction can be realized.

The life prediction module 11 may be one or include a plurality of processors, and the life prediction module 11 is provided in a device (e.g., a vehicle) using keys. The life prediction module 11 may receive the service life prediction model generated by the model generation module 12, and predict the remaining service life of the battery of the life prediction module 11 based on the service life prediction model and the real-time data collected by the life prediction module 11. The service life prediction model is constructed based on experimental data and real data collected by equipment of the same type as the service life prediction module 11.

The model generation module 12 may be one or include multiple processors. The model generation module 12 may acquire real data acquired by the life prediction module 11 from the life prediction module 11, may acquire experimental data from the data storage module 13, and may generate the above-described service life prediction model based on the real data and the experimental data, and then upload the service life prediction model to the life prediction module 11. In some embodiments, the model generation module 12 may also upload the life prediction model to the data storage module 13.

The data storage module 13 is used for storing the experimental data. In some embodiments, the data storage module 13 may also be used to store upper life prediction models.

In one embodiment, the life prediction module 11, the model generation module 12, and the data storage module 13 are three independent modules, and in this embodiment, the life prediction module 11 is provided in a device (e.g., a vehicle) using a key.

In other embodiments, the life prediction module 11, the model generation module 12, and the data storage module 13 may be integrated into a device that is a device that uses keys, such as a vehicle.

Before introducing the methods provided in the embodiments of the present application, the following terms and their corresponding examples will be explained. Referring to table 1, the left side of table 1 is the following Chinese nouns, and the right side is an example corresponding to the Chinese nouns.

TABLE 1

Chinese noun	Examples of the invention
		Real data	Voltage_real，X1，X2，X3，X4，…，Xn
Data detected by voltage sensor	Voltage_real
		Data detected by a rainfall sensor	X1
Data detected by the temperature sensor	X2
		Number of times of opening and closing door	X3
Number of key presses	X4
		Other data related to the target data	Xn
Object data	Voltage_real
		Target sample data	Voltage_sample
Remaining service life of battery	Lifetime
		Current remaining useful life of the battery	Lifetime_present
Experimental model	f(Voltage_sample,Lifetime)
		Relation of objective function	f(Voltage_real,Lifetime)
Error distribution	Error＝f(Voltage_sample,Voltage_real)
		Target regression prediction model	f(Voltage_real，X1，X2，X3，X4，…，Xn)
Service life prediction model	f(Lifetime，X1，X2，X3，X4，…，Xn)
		Current true data	X1'，X2'，X3'，X4'，…，Xn'
Data after current real data preprocessing	X1”，X2”，X3”，X4”，…，Xn”

Referring to fig. 2, fig. 2 is a schematic flowchart illustrating a method for generating a battery life prediction model according to an embodiment of the present disclosure. The method for generating the battery service life prediction model may be applied to the model generation module 12, or the battery service life prediction model generation apparatus 500 shown in fig. 7, which will be mentioned later, or the vehicle 700 shown in fig. 9, which will be mentioned later. The service life prediction model may include the following steps S110 to S140.

Step S110, acquiring real data and experimental sample data of preset types.

The preset type of real data is data acquired by the equipment in real time. The preset type of real data may be data related to the service life of the battery. In some embodiments, the preset type of real data may include one or a combination of data detected by a voltage sensor, data detected by a rainfall sensor, data detected by a temperature sensor, the number of times the door of the device is opened and closed, and the number of times the key is pressed. It should be noted that the preset type of actual data may be determined according to actual requirements, and may further include other data not listed herein, which is related to the service life of the battery, and the embodiment of the present application is not specifically limited herein.

The experimental sample data is original experimental data provided by a manufacturer or an experimental model established based on the original experimental data. As an example, the raw experimental data includes data collected while controlling internal resistance, current, voltage, and temperature of the battery. As another example, the raw experimental data may be historical data of previous similar devices or the same battery model, and may be, but is not limited to, voltage detected by a voltage sensor, temperature detected by a temperature sensor, rainfall detected by a rainfall sensor, number of key presses, number of door opens and closes of the device. The experimental model is a manufacturer-provided sample model that includes a functional relationship between the target sample data and the remaining service life of the battery in step S120.

It should be noted that, before the experimental model is established, the raw experimental data needs to be cleaned and preprocessed. The cleaning and preprocessing refer to the steps of removing abnormal values in the sample data of the laboratory by adopting correlation analysis, heat map and other modes, filling missing values in the sample data of the laboratory, and screening the sample data of the laboratory with high correlation.

In some embodiments, the device continuously collects and stores real data of a preset type in a first storage area of the device during operation. The real data of the preset type can be directly acquired from the first storage area of the device. The device receives and stores the experimental sample data uploaded to the device by the manufacturer in advance into a second storage area of the device. The experimental sample data can be directly acquired from the second storage area. The first storage area and the second storage area may be located in the same storage area or two storage areas that are independent of each other, and embodiments of the present application are not limited in this respect.

And step S120, determining a functional relation between the target sample data and the residual service life of the battery according to the experimental sample data.

The target sample data refers to data in the experiment sample data, which has a strong correlation with real data of the same type, and as an example, the target sample data is an experiment voltage.

In other embodiments, the experimental sample data is the experimental model, and the experimental model may be directly used as a functional relationship between the target sample data and the remaining service life of the battery.

In some embodiments, the experimental sample data is original experimental data provided by a manufacturer, and the original experimental data may be cleaned and preprocessed, and the experimental model may be established according to the experimental data after cleaning and preprocessing, and the experimental model may be used as a functional relationship between the target sample data and the remaining service life of the battery.

Step S130, determining target data from the real data, wherein the target data and the target sample data belong to the same type of data.

The target data refers to data which has strong relevance with the same type of experimental sample data in the real data, and as an example, the target data is the voltage of a battery acquired by the equipment in real time through a voltage sensor.

In some embodiments, target sample data may be obtained, and the target data may be determined from the real data according to the target sample data. As an example, the target sample data is an experimental voltage, and the target data is a voltage of a battery acquired by the device in real time through a voltage sensor.

Before determining the target data from the real data, the real data needs to be cleaned and preprocessed. Here, the operation of cleaning and preprocessing is similar to the above-mentioned operation of cleaning and preprocessing the original experimental data, please refer to relevant parts in step S110, and the embodiments of the present application are not described herein again.

It should be noted that, the sequence of step S120 and step S130 may be exchanged, that is, the functional relationship between the target sample data and the remaining service life of the battery may be determined first, or the target data may be determined first, and the embodiment of the present application is not limited specifically herein.

And S140, generating a service life prediction model according to the real data, the target sample data and the functional relation (experimental model), wherein the service life prediction model is used for predicting the residual service life of the battery.

The service life prediction model comprises a functional relation between other data except the target data in the real data and the remaining service life of the battery.

In some embodiments, the functional relationship between the other data than the target data in the real data and the remaining service life of the battery may be determined according to the real data, the target sample data and the functional relationship, and the service life prediction model may be established based on the functional relationship between the other data than the target data in the real data and the remaining service life of the battery.

For detailed description of step S140, refer to steps S240 to S270 below.

According to the method for generating the battery service life prediction model, the preset type of real data and the experiment sample data are adopted to generate the service life prediction model, the relation between the experiment data and the real data is considered while the residual service life of the battery is predicted through a large amount of abundant data, the generated service life prediction model has abundant data bases and is strong in relevance with the real data, and therefore the accuracy of the residual service life prediction model on the battery is improved. Meanwhile, the service life prediction model can accurately analyze the residual capacity of the key by determining the target function relationship between target data (real-time voltage) and the residual service life of the battery.

Referring to fig. 3, fig. 3 is a method for generating a battery life prediction model according to another embodiment of the present disclosure. The method for generating the battery service life prediction model may be applied to the model generation module 12, or the battery service life prediction model generation apparatus 500 shown in fig. 7, which will be mentioned later, or the vehicle 700 shown in fig. 9, which will be mentioned later. The service life prediction model may include the following steps 210 to S270.

Step S210, acquiring real data and experimental sample data of preset types,

step S220, according to the experimental sample data, determining the functional relationship between the target sample data and the residual service life of the battery.

Step S230, determining target data from the real data, where the target data and the target sample data belong to the same type of data.

Please refer to the above steps S110 to S130 in steps S210 to S230, which are not described herein again in this embodiment of the present application.

Step S240, performing error analysis on the target data and the target sample data to obtain error distribution between the target data and the target sample data.

The error distribution refers to data for representing the inconsistency between the target data and the target sample data.

In some embodiments, the target data and the target sample data may be compared and analyzed, for example, whether the degrees of change between the target data and the target sample data are similar or not is determined, so as to determine an error between the target data and the target sample data, and obtain an error distribution between the target data and the target sample data.

In some embodiments, if the error distribution of the target data and the target sample data does not satisfy the standard normal distribution, the target data and the target sample data may be subjected to a generalized power change, and the data after the generalized power change may be verified to confirm the error distribution. Validation herein may include, but is not limited to, disorder detection and diffuse distribution. Wherein the disorder detection can confirm whether the error distributions are not obviously related. The dispersion distribution can confirm whether the error distribution is uniformly distributed in each region.

In other embodiments, if the error between the target data and the target sample data is large, the generalized power change may be performed on the target data and the target sample data, and the verification may be performed on the data after the generalized power change, so as to confirm the error distribution.

And step S250, determining the objective function relationship between the objective data and the remaining service life of the battery according to the error distribution and the functional relationship.

The functional relationship can represent the relationship between target sample data and the residual service life of the battery, and the error distribution can represent the relationship between the target data and the target sample data. The functional relation and the error distribution both comprise target sample data, so that the target functional relation between the target data and the residual service life of the battery can be determined according to the error distribution and the functional relation. Specifically, a mapping relationship between the target data and the target sample data may be determined according to the error distribution, and the target data is used to replace the target sample data in the functional relationship based on the mapping relationship, so as to obtain a target functional relationship between the target data and the remaining service life of the battery.

As an example, the Error distribution is Error = f (Voltage _ sample, voltage _ real), the function relationship is f (Voltage _ sample, lifetime), and the objective function relationship is f (Voltage _ real, lifetime) can be obtained according to the Error distribution and the function relationship.

Step S260, a target regression prediction model between the target data and the real data is established according to the target data and the real data.

All data of the real data have correlation, that is, other data in the real data except the target data are all correlated with the target data.

The target regression prediction model comprises a functional relation between target data and other data of real data except the target data. The target regression prediction model may include, but is not limited to, one or more combinations of a linear regression prediction model, a logistic regression prediction model, a ridge regression prediction model, a stepwise regression prediction model, a lasso regression prediction model, and an elastic regression prediction model.

The functional relationship between the target data and the other data of the real data except the target data can be determined according to the correlation between the target data and the real data, and the target regression prediction model between the target data and the real data is established based on the functional relationship between the target data and the other data of the real data except the target data.

For example, if the target data is denoted by Voltage _ real, and the real data is denoted by Voltage _ real, X1, X2, X3, X4, \8230, xn, then the target regression prediction model is f (Voltage _ real, X1, X2, X3, X4, \8230, xn).

In some embodiments, step S260 may specifically include the following steps: according to the target data and the real data, preliminarily establishing a preliminary regression prediction model between the target data and the real data; adjusting parameters of the preliminary regression prediction model by adopting a random Taylor expansion of a Lagrange component to obtain an adjusted regression prediction model; and verifying the adjusted regression prediction model to obtain a target regression prediction model.

Specifically, a plurality of preliminary prediction regression models may be obtained by respectively establishing a preliminary regression prediction model corresponding to each type of data, except for the target data, in the real data and the target data. And adjusting the parameters of each preliminary regression prediction model by adopting a random Taylor expansion of a Lagrange component to obtain the optimal parameters of each preliminary regression prediction model. And determining an optimal parameter combination according to each optimal parameter, and combining a plurality of preliminary regression prediction models corresponding to the optimal parameter combination to obtain the adjusted regression prediction model.

Each of the plurality of preliminary prediction regression models refers to a model that is created based on a correlation between target data and data of the same type other than the target data in the real data. The multiple preliminary prediction regression models may be the same type of regression model, or may include multiple types of regression models, and the embodiments of the present application are not limited in this respect. For ease of calculation, the plurality of preliminary predictive regression models may be selected to be the same type of regression model. The regression model may include, but is not limited to, one or more combinations of a linear regression prediction model, a logistic regression prediction model, a ridge regression prediction model, a stepwise regression prediction model, a lasso regression prediction model, and an elastic regression prediction model.

As an example, the plurality of preliminary prediction regression models are represented as f1 (Voltage _ real, X1), f2 (Voltage _ real, X2), f3 (Voltage _ real, X3), f4 (Voltage _ real, X4), \8230;, fn (Voltage _ real, xn), and the regression prediction model after adjustment may be f (Voltage _ real, X1, X2, X3, X4, \8230;, xn).

In some embodiments, adjusting the parameters of each preliminary regression prediction model by using a random taylor expansion of lagrange's polynomials to obtain the optimal parameters of each preliminary regression prediction model may include the following steps:

when the average of the same type of data (for example, data corresponding to all the gate opening times) in the real data is zero and the variance is a finite number, the presence parameter λ satisfies the following expression:

wherein n represents the number of data, j represents the jth data, and X _j The characteristics of the jth X are characterized,

is an estimated value, characterizing the nth parameter.

By adopting taylor random expansion formula for the above expression (1), the expression of λ can be obtained as follows:

where D denotes the convergence of the distribution, σ ² And characterizing the variance of the same type of data in the real data, wherein N characterizes N types of data in the real data.

According to the expressions (1) and (2), an optimal solution of λ can be calculated, and parameters of the preliminary regression prediction model can be adjusted according to the optimal solution.

In some embodiments, the verification method for verifying the regression prediction model after adjustment may include, but is not limited to, a cross-validation method and a regression equation verification method, which may include, but is not limited to, a joint hypothesis test, F-test, and T-test.

And step S270, generating a service life prediction model according to the target function relation and the target regression prediction model.

In some embodiments, step S270 may include the steps of: and performing model fusion on the target function relation and the target regression prediction model by adopting a preset model fusion method to obtain a service life prediction model. The preset model fusion method may select an existing model fusion algorithm according to actual requirements, for example, the preset model fusion method may include, but is not limited to, a Bagging algorithm (Bagging algorithm, also referred to as a guided aggregation algorithm and a Bagging algorithm) and a Boosting algorithm (Boosting algorithm).

Referring to fig. 4, fig. 4 is a schematic flowchart illustrating a method for predicting remaining useful life of a battery according to an embodiment of the present disclosure. The method for predicting the remaining useful life of the battery is applied to the above-described life prediction module 11, or the device 600 for predicting the remaining useful life of the battery shown in fig. 8, which will be mentioned later, or the vehicle 700 shown in fig. 9, which will be mentioned later. The method for predicting the remaining useful life of the battery may include the following steps S310 and S320.

Step S310, obtaining current real data of a preset type.

And the current real data of the preset type is real data acquired by current equipment in real time. The preset type of current real data refers to data of the same type as the variables in the service life prediction model. The preset type of current real data includes other data than the target data among the preset type of real data.

As an example, the target data in the preset type of real data is a voltage, and the preset type of current real data includes data other than the voltage in the preset type of real data, for example, the preset type of current real data includes one or a combination of several of data detected by a current rainfall sensor, data detected by a current temperature sensor, the number of times of opening and closing a door of the current device, and the number of times of pressing a key of the current device.

The steps of the method for obtaining the current real data of the preset type are similar to the steps of the method for obtaining the real data of the preset type, so the detailed description of the step S310 please refer to the step S110, which is not described herein again.

Step S320, predicting the remaining service life of the current battery based on the service life prediction model established according to the method for generating the battery service life prediction model.

In some embodiments, the current real data may need to be preprocessed before predicting the remaining useful life of the current battery. Preprocessing includes, but is not limited to, interference elimination, missing value filling, time-dependent variable conversion, and generalized power variation to ensure that the preprocessed data meet the requirements of the service life prediction model and that the variables of the preprocessed data are consistent with the parameter properties of the same variables in the service life prediction model, thereby improving the accuracy of the prediction result.

In some embodiments, the data after the preprocessing may be input into the service life prediction model, and the output of the service life prediction model may be used as the remaining service life of the current battery. The remaining life of current battery can reflect the remaining capacity of battery, consequently, through the remaining life of current battery, can accurate analysis key's remaining capacity.

According to the method for predicting the remaining service life of the battery, the remaining service life of the current battery is predicted according to the service life prediction model, the accuracy of a prediction result can be improved, and troubles caused by the fact that a sudden key is not electrified to a user are avoided. Meanwhile, the service life prediction model can accurately analyze the residual capacity of the key by determining the target function relationship between the target data (real-time voltage) and the residual service life of the battery.

Referring to fig. 5, fig. 5 is a schematic flowchart illustrating a method for predicting remaining useful life of a battery according to another embodiment of the present disclosure. The method for predicting the remaining useful life of the battery is applied to the above-described life prediction module 11, or the device 600 for predicting the remaining useful life of the battery shown in fig. 8, which will be mentioned later, or the vehicle 700 shown in fig. 9, which will be mentioned later. The method for predicting the remaining service life of the battery may include the following steps S410 to S430.

In step S410, the current real data of the preset type is obtained.

Step S420, predicting the remaining service life of the current battery based on the service life prediction model established according to the method for generating the battery service life prediction model.

Please refer to step S310 and step S320 for step S410 and step S420, respectively, which are not described herein again in this embodiment of the present application.

Step S430, if the remaining service life of the current battery is lower than the preset threshold, sending a prompt message for replacing the battery or displaying the remaining service life of the current battery.

The preset threshold may be set by a user according to an actual requirement, for example, the preset threshold may be 5%, and the embodiment of the present application is not specifically limited herein.

In some embodiments, after predicting the remaining service life of the current battery, it may be detected whether the remaining service life of the current battery is lower than a preset threshold.

If the remaining service life of the current battery is lower than the preset threshold value, a prompt message for replacing the battery can be sent out or the remaining service life of the current battery can be displayed. The manner of sending the prompt information for replacing the battery and displaying the remaining service life of the current battery may be a manner of sending a voice through sound equipment of the device, or a manner of displaying a text, an image or a video through a display screen of the device, or a combination of the two manners, which is not specifically limited herein.

In some embodiments, if the remaining service life of the current battery is not lower than the preset threshold, the execution of steps S410 to S430 may be ended, and the execution of steps S410 to S430 may be restarted when the prediction instruction is received. Wherein the prediction instruction is input by the user by voice, or by pressing or clicking a key on a display screen of the device, or by pressing a button on the device for initiating the battery remaining life prediction operation. The prediction instruction is used to start execution of operations of step S410 to step S430.

In other embodiments, if the remaining service life of the current battery is not lower than the preset threshold, the steps S410 to S430 may be executed again after a preset time period. The preset time period may be set according to actual needs, for example, the preset time period may be one week, and the embodiment of the present application is not specifically limited herein.

According to the method for predicting the remaining service life of the battery, the remaining service life of the current battery is predicted according to the service life prediction model, the accuracy of a prediction result can be improved, and troubles caused by the fact that a sudden key is not electrified to a user are avoided. Meanwhile, the service life prediction model can accurately analyze the residual electric quantity of the key by determining the target function relation between the target data (real-time voltage) and the residual service life of the battery, and remind a user to replace the battery at a proper time, so that the waste of resources and cost caused by frequent replacement of the battery by the user can be reduced.

It should be noted that, if the main executing bodies of the method for generating the battery service life prediction model and the method for predicting the remaining service life of the battery are located in the same device, the two methods may be combined into one set of method, that is, the method for predicting the remaining service life of the battery. Please refer to relevant parts in the foregoing embodiments for a specific implementation of the method for predicting the remaining service life of a battery, which is not described herein again in this application.

For convenience of understanding, the embodiments of the present application provide an exemplary embodiment, which is only used for illustrating the methods provided by the embodiments of the present application and should not be construed as limiting the methods provided by the embodiments of the present application. Referring to fig. 6, fig. 6 is a schematic flowchart illustrating a method for predicting a remaining service life of a battery according to an exemplary embodiment of the present disclosure. In the present exemplary embodiment, the method of predicting the remaining useful life of the battery is applied to a vehicle in which the life prediction module 11 and the model generation module 12 are integrated.

As shown in fig. 6, the method for predicting the remaining service life of the battery includes a method for generating a life prediction model and a method for predicting the remaining service life of the battery, and specifically includes the following steps:

step S1, real vehicle data Voltage _ real, X1, X2, X3, X4, \ 8230;, xn are obtained.

Wherein the real vehicle data is the real data of the preset type.

And S2, cleaning and preprocessing the real vehicle data to obtain cleaned and preprocessed data.

And S3, acquiring target data Voltage _ real from the cleaned and preprocessed data in the step S2.

And S4, screening data X1, X2, X3, X4, \ 8230and Xn associated with the target data Voltage _ real in the real vehicle data.

And S5, establishing a target regression prediction model f (Voltage _ real, X1, X2, X3, X4, \ 8230;, xn) according to the target data Voltage _ real and the data X1, X2, X3, X4, \ 8230;, xn associated with the target data.

The target regression prediction model can be established by executing the steps S1 to S5.

And S6, acquiring target sample data Voltage _ sample and the remaining service life of the battery, namely Life.

And S7, cleaning and preprocessing the target sample data, namely, the Voltage _ sample and the remaining service life of the battery, to obtain cleaned and preprocessed data.

Step S8, establishing an experimental model f (Voltage _ sample, lifetime) according to the data after cleaning and preprocessing in step S7.

Step S9, detecting whether the experimental model f (Voltage _ sample, lifetime) reaches the expectation.

If the experimental model f (Voltage _ sample, lifetime) reaches the expectation, the step S10 is continuously executed, and if it is detected that the experimental model f (Voltage _ sample, lifetime) does not reach the expectation, the step S8 is returned to be executed.

In some embodiments, whether the experimental model meets the expectation may be detected, and whether the parameter of the target regression prediction model meets a preset parameter threshold may be set by a developer according to experience, or according to a requirement for accuracy of a prediction result, which is not specifically limited herein.

Step S10, outputting the experimental model f (Voltage _ sample, lifetime).

Step S11, according to the target data Voltage _ real in step S3 and the target sample data Voltage _ sample in step S6, screening a relationship (i.e. the above-mentioned Error distribution) Error = f (Voltage _ sample, voltage _ real) between the target data Voltage _ real and the target sample data Voltage _ sample.

Step S12, establishing a relationship (i.e. the above-mentioned objective function relationship) f (Voltage _ real, lifetime) between the target data and the remaining service life of the battery according to the above-mentioned relationship Error = f (Voltage _ sample, voltage _ real) and the experimental model f (Voltage _ sample, lifetime).

By executing steps S6 to S12, a relationship f (Voltage _ real, lifetime) between the target data and the remaining service life of the battery can be established.

And step S13, establishing a service life prediction model f (Lifetime, X1, X2, X3, X4, 8230, xn) according to the target regression prediction model f (Voltage _ real, X1, X2, X3, X4, \ 8230;, xn) in the step S5 and the relation f (Voltage _ real, lifetime) in the step S12.

Step S14, detecting whether the service life prediction model f (Lifetime, X1, X2, X3, X4, \ 8230;, xn) reaches the expectation.

If the service life prediction model f (Lifetime, X1, X2, X3, X4, \8230;, xn) reaches the expectation, the step S15 is continuously executed, and if the service life prediction model f (Lifetime, X1, X2, X3, X4, \8230;, xn) does not reach the expectation, the step S13 is executed again.

In some embodiments, detecting whether the service life prediction model reaches the expectation may detect whether a parameter of the service life prediction model meets a preset parameter threshold, where the preset parameter threshold may be set by a developer according to experience or according to a requirement for accuracy of a prediction result, and the embodiment of the present application is not limited specifically herein.

And S15, outputting the service life prediction model f (Lifetime, X1, X2, X3, X4, \8230;, xn) to a vehicle end.

Step S16, the vehicle end collects real-time data (i.e. the current real data mentioned above) X1', X2', X3', X4', \ 8230;, xn '.

Step S17, preprocessing the real-time data X1', X2', X3', X4', \8230andXn 'to obtain preprocessed data X1', X2', X3', X4', \8230andXn'.

And S18, predicting the residual service life of the current battery by adopting a service life prediction model f (Lifetime, X1, X2, X3, X4, \ 8230;, xn) according to the preprocessed data X1', X2', X3', X4', \ 8230;, xn in the step S17 to obtain the residual service life Lifetime _ present of the current battery.

And S19, when the remaining service life of the current battery, namely the Lifetime _ present, is lower than a preset threshold value, sending an early warning to a user.

The preset threshold may be set according to an actual requirement, for example, the preset threshold is 5%, and the embodiment of the present application is not specifically limited herein.

The manner of sending the warning to the user is the same as or similar to the manner of sending the prompt information for replacing the battery or displaying the remaining service life of the current battery, which is described above, and reference is specifically made to the detailed description of step S430, which is not repeated herein in this embodiment of the present application.

In some embodiments, when the remaining service life of the current battery is not lower than the preset threshold, the execution of steps S16 to S19 may be ended, and the execution of steps S16 to S19 may be restarted when the prediction instruction is received. Wherein the prediction instruction is input by a user through voice, or by pressing or clicking a key on a display screen of the device, or by pressing a button on the device for initiating the battery remaining life prediction operation. The prediction instruction is used to start execution of the operations of step S16 to step S19.

In other embodiments, when the remaining service life of the current battery is not lower than the preset threshold, steps S16 to S19 may be executed again after a preset time period. The preset time period may be set according to actual needs, for example, the preset time period may be one week, and the embodiment of the present application is not specifically limited herein.

It should be noted that, for portions that are not described in detail in steps S1 to S19, reference is made to relevant portions of the foregoing embodiments, and details of the embodiments of the present application are not repeated herein.

Referring to fig. 7, fig. 7 is a block diagram illustrating a structure of a device for generating a battery life prediction model according to an embodiment of the present disclosure. The generation device 500 of the battery service life prediction model may be applied to the above-described life prediction module 11 or a vehicle 700 shown in fig. 9, which will be mentioned later.

The generation device 500 of the battery service life prediction model comprises a data acquisition module 510, a function determination module 520, a data extraction module 530 and a model generation module 540. Wherein:

the data obtaining module 510 is configured to obtain real data and experimental sample data of a preset type.

And a function determining module 520, configured to determine a functional relationship between the target sample data and the remaining service life of the battery according to the experiment sample data.

The data extraction module 530 is configured to determine target data from the real data, where the target data and the target sample data belong to the same type of data.

The model generating module 540 is configured to generate a service life prediction model according to the real data, the target sample data, and the functional relationship, where the service life prediction model is used to predict the remaining service life of the battery.

In some embodiments, the model generation module 540 includes an error analysis sub-module, a relationship analysis sub-module, and a model generation sub-module. Wherein:

and the error analysis submodule is used for carrying out error analysis on the target data and the target sample data to obtain the error distribution between the target data and the target sample data.

And the relation analysis submodule is used for determining a target function relation between the target data and the residual service life of the battery according to the error distribution and the function relation.

And the model generation submodule is used for generating the service life prediction model according to the objective function relation, the objective data and the real data.

In some embodiments, the model generation submodule includes a first model generation unit and a second model generation unit. Wherein:

and the first model generation unit is used for establishing a target regression prediction model between the target data and the real data according to the target data and the real data.

And the second model generation unit is used for generating the service life prediction model according to the objective function relation and the objective regression prediction model.

In some embodiments, the first model generation unit comprises a model preliminary generation subunit, a parameter adjustment subunit, and a model generation subunit. Wherein:

and the model preliminary generation subunit is used for preliminarily establishing a preliminary regression prediction model between the target data and the real data according to the target data and the real data.

And the parameter adjusting subunit is used for adjusting the parameters of the preliminary regression prediction model by adopting a random Taylor expansion of a Lagrange component to obtain the adjusted regression prediction model.

And the model generation subunit is used for verifying the adjusted regression prediction model to obtain the target regression prediction model.

Referring to fig. 8, fig. 8 is a block diagram illustrating a device for predicting remaining battery life according to an embodiment of the present disclosure. The prediction 600 of the remaining useful life of the battery may be applied to the model generation module 12 described above or to a vehicle 700 shown in fig. 9 to be mentioned later.

The prediction 600 of remaining useful life of the battery includes a data acquisition module 610 and a life prediction module 620. Wherein:

the data obtaining module 610 is configured to obtain current real data of a preset type.

And the service life prediction module 620 is configured to predict the remaining service life of the current battery based on the service life prediction model established according to the method for generating the battery service life prediction model.

In some embodiments, the prediction 600 of remaining useful life of the battery further includes a reminder module.

And the reminding module is used for sending out prompt information for replacing the battery or displaying the remaining service life of the current battery if the remaining service life of the current battery is lower than a preset threshold value.

It can be clearly understood by those skilled in the art that the device 500 for generating a battery service life prediction model provided in the embodiment of the present application can implement the method for generating a battery service life prediction model provided in the embodiment of the present application, and the device 600 for predicting the remaining service life of a battery provided in the embodiment of the present application can implement the method for predicting the remaining service life of a battery provided in the embodiment of the present application. The specific working processes of the above devices and modules may refer to corresponding processes in the embodiments of the method of the present application, and are not described herein again.

In the embodiments provided in this application, the coupling, direct coupling or communication connection between the modules shown or discussed may be an indirect coupling or communication coupling through some interfaces, devices or modules, and may be in an electrical, mechanical or other form, which is not limited in this application.

In addition, each functional module in the embodiments of the present application may be integrated into one processing module, or each module may exist alone physically, or two or more modules are integrated into one module. The integrated module may be implemented in the form of hardware, or may be implemented in the form of a functional module of software, and the embodiment of the present application is not limited herein.

Referring to fig. 9, fig. 9 is a structural block diagram of a vehicle according to an embodiment of the present application. The vehicle 700 may include one or more of the following components: memory 710, one or more processors 720, and one or more applications, wherein the one or more applications may be stored in the memory 710 and configured to cause the one or more processors 720 to perform the methods provided by the embodiments of the present application when invoked by the one or more processors 720.

Processor 720 may include one or more processing cores. The processor 720 interfaces with various interfaces and circuitry throughout the vehicle 700 for executing or executing instructions, programs, code sets, or instruction sets stored in the memory 710, as well as invoking execution or execution of data stored in the memory 710, performing various functions of the vehicle 700 and processing the data. Alternatively, the processor 720 may be implemented in at least one hardware form of Digital Signal Processing (DSP), field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). Processor 720 may integrate one or a combination of a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), and a modem. Wherein, the CPU mainly processes an operating system, a user interface, an application program and the like; the GPU is used for rendering and drawing display content; the modem is used to handle wireless communications. It is understood that the modem may not be integrated into the processor 720, but may be implemented by a communication chip.

The Memory 710 may include a Random Access Memory (RAM) or a Read-Only Memory (ROM). The memory 710 may be used to store instructions, programs, code sets, or instruction sets. The memory 710 may include a program storage area and a data storage area. Wherein the stored program area may store instructions for implementing an operating system, instructions for implementing at least one function, instructions for implementing the various method embodiments described above, and the like. The storage data area may store data created by the vehicle 700 in use, and the like.

Referring to fig. 10, fig. 10 is a block diagram of a computer readable storage medium according to an embodiment of the present disclosure. The computer-readable storage medium 800 has stored therein a program code 810, the program code 810 being configured to, when invoked by a processor, cause the processor to perform the methods provided by embodiments of the present application.

The computer-readable storage medium 800 may be an electronic Memory such as a flash Memory, an Electrically-Erasable Programmable Read-Only-Memory (EEPROM), an Erasable Programmable Read-Only-Memory (EPROM), a hard disk, or a ROM. Optionally, the Computer-Readable Storage Medium 800 includes a Non-volatile Computer-Readable Medium (Non-TCRSM). The computer readable storage medium 800 has storage space for program code 810 for performing any of the method steps described above. The program code 810 can be read from or written to one or more computer program products. The program code 810 may be compressed in a suitable form.

In summary, the embodiment of the application provides a method and a device for generating a battery service life prediction model, and a vehicle, the method and the device adopt preset type real data and experiment sample data to generate the service life prediction model, and consider the relation between the experiment data and the real data while predicting the remaining service life of a battery through a large amount of abundant data, so that the generated service life prediction model has abundant data bases and strong relevance with the real data, thereby improving the accuracy of the service life prediction model in predicting the remaining service life of the battery.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present application, and are not limited thereto. Although the present application has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

1. A method for generating a battery service life prediction model is characterized by comprising the following steps:

acquiring real data and experimental sample data of a preset type;

determining a functional relation between target sample data and the residual service life of the battery according to the experimental sample data;

determining target data from the real data, wherein the target data and the target sample data belong to the same type of data;

and generating a service life prediction model according to the real data, the target sample data and the functional relation, wherein the service life prediction model is used for predicting the residual service life of the battery.

2. The method of claim 1, wherein generating a service life prediction model based on the real data, the target sample data, and the functional relationship comprises:

performing error analysis on the target data and the target sample data to obtain error distribution between the target data and the target sample data;

determining a target functional relationship between the target data and the remaining service life of the battery according to the error distribution and the functional relationship;

and generating the service life prediction model according to the objective function relation, the objective data and the real data.

3. The method of claim 2, wherein generating the service life prediction model from the objective functional relationship, the objective data, and the real data comprises:

establishing a target regression prediction model between the target data and the real data according to the target data and the real data;

and generating the service life prediction model according to the target function relation and the target regression prediction model.

4. The method of claim 3, wherein the building a target regression prediction model between the target data and the real data from the target data and the real data comprises:

according to the target data and the real data, preliminarily establishing a preliminary regression prediction model between the target data and the real data;

adjusting parameters of the preliminary regression prediction model by adopting a random Taylor expansion of a Lagrange's element to obtain an adjusted regression prediction model;

and verifying the adjusted regression prediction model to obtain the target regression prediction model.

5. The method according to any one of claims 1 to 4, wherein the predetermined type of authenticity data comprises one or more of data detected by a voltage sensor, data detected by a rain sensor, data detected by a temperature sensor, the number of door opening and closing times of the device, and the number of key pressing times.

6. A method for predicting remaining service life of a battery, comprising:

acquiring current real data of a preset type;

predicting the remaining service life of the current battery based on a service life prediction model established according to the method of any one of claims 1 to 5.

7. The method of claim 6, further comprising, after said predicting the remaining useful life of the current battery:

and if the remaining service life of the current battery is lower than a preset threshold value, sending out prompt information for replacing the battery or displaying the remaining service life of the current battery.

8. An apparatus for generating a battery service life prediction model, comprising:

the data acquisition module is used for acquiring real data and experimental sample data of a preset type, wherein the real data is data acquired by equipment;

the function determining module is used for determining the functional relationship between the target sample data and the residual service life of the battery according to the experimental sample data;

the data extraction module is used for determining target data from the real data, and the target data and the target sample data belong to the same type of data;

and the model generation module is used for generating a service life prediction model according to the real data, the target sample data and the functional relation, and the service life prediction model is used for predicting the residual service life of the battery.

9. A vehicle, characterized by comprising:

a memory;

one or more processors;

one or more applications, wherein the one or more applications are stored in the memory and configured to, when invoked by the one or more processors, cause the one or more processors to perform the method of any of claims 1-7.