Method, device and equipment for constructing elastic network regression model of battery retention rate

Abstract

The invention relates to a method, a device and equipment for constructing an elastic network regression model of a battery retention rate. The method comprises the following steps: acquiring sample data of a sample battery and an initial elastic network regression model; inputting the sample data into the initial elastic network regression model, and outputting a prediction result; determining the target loss of the initial elastic network regression model according to the prediction result; the target loss comprises an error term and a regularization term; and training the initial elastic network regression model according to the target loss to obtain a trained elastic network regression model. According to the method, penalty for the regression coefficient is introduced into the regularization item, the complexity of the elastic network regression model is controlled, and the elastic network regression model is prevented from being over-fitted, so that the training efficiency of the initial elastic network regression model is improved, and due to the fact that the number of sample data needed for training the regression model is small, the training efficiency of the regression model is improved. And the complexity of training the initial elastic network regression model is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of lithium-ion batteries, and in particular to a method, device and equipment for constructing an elastic network regression model of battery retention rate. Background Art

[0002] Lithium-ion batteries have the characteristics of high energy density, long cycle life and environmental friendliness. They are widely used in portable electronic devices, electric vehicles and energy storage systems. However, as the use time of lithium-ion batteries increases, the capacity of lithium-ion batteries will gradually decay, leading to the termination of battery use. Therefore, predicting the capacity retention rate of lithium-ion batteries is of great significance for evaluating battery performance, guiding battery design, optimizing battery management, discovering potential problems and reducing development costs and time.

[0003] Typically, a machine learning model is trained by collecting data from the entire life cycle of lithium-ion batteries, and the trained machine learning model is used to predict the capacity retention rate of lithium-ion batteries.

[0004] However, conventional technologies have a problem of low prediction efficiency in predicting the capacity retention rate of lithium-ion batteries. Summary of the invention

[0005] Based on this, it is necessary to provide a method, device and equipment for constructing an elastic network regression model of battery retention rate that can improve the prediction efficiency in response to the above technical problems.

[0006] In a first aspect, the present application provides a method for constructing an elastic network regression model of battery retention rate, comprising:

[0007] Obtain sample data and an initial elastic network regression model of a sample battery;

[0008] Inputting the sample data into the initial elastic network regression model and outputting a prediction result;

[0009] Determine the target loss of the initial elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term;

[0010] The initial elastic network regression model is trained according to the target loss to obtain a trained elastic network regression model.

[0011] In one embodiment, obtaining sample data of a sample battery includes:

[0012] Acquiring initial data of the sample battery;

[0013] Performing correlation screening on a plurality of initial features in the initial data to determine a plurality of intermediate features of the sample battery;

[0014] The intermediate features are screened according to the multicollinearity of the intermediate features to determine the sample data.

[0015] In one embodiment, screening the intermediate features according to the multicollinearity of the intermediate features to determine the sample data includes:

[0016] Determining a variance inflation factor for each of the intermediate features;

[0017] The sample data is determined according to the variance inflation factor of each of the intermediate features and the intermediate features of a first preset threshold.

[0018] In one embodiment, determining the variance inflation factor of each of the intermediate features includes:

[0019] Determine the data fluctuation amount and data variation amount of each intermediate feature according to the average value of each intermediate feature and each intermediate feature;

[0020] Determining the coefficient of determination of each intermediate feature according to the data fluctuation amount and the data variation amount of each intermediate feature;

[0021] The variance inflation factor of each intermediate feature is determined according to the square value of the determination coefficient of each intermediate feature.

[0022] In one embodiment, the step of performing correlation screening on the multiple initial features in the initial data to determine the multiple intermediate features of the sample battery includes:

[0023] Determining the Pearson correlation coefficient between each of the initial characteristics and the capacity retention rate;

[0024] The initial feature whose Pearson correlation coefficient is greater than a second preset threshold is determined as the intermediate feature.

[0025] In one embodiment, the method further comprises:

[0026] Acquiring characteristic information of a battery to be predicted, and performing standardization processing on the characteristic information of the battery to be predicted to obtain processed characteristic information;

[0027] The processed feature information is input into the trained elastic network regression model to determine the capacity retention rate of the battery to be predicted.

[0028] In a second aspect, the present application also provides a device for constructing an elastic network regression model of a battery retention rate, comprising:

[0029] An acquisition module, used to acquire sample data of a sample battery and an initial elastic network regression model;

[0030] A first prediction module, used for inputting the sample data into the initial elastic network regression model and outputting a prediction result;

[0031] A determination module, used to determine the target loss of the elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term;

[0032] A training module is used to train the initial elastic network regression model according to the target loss to obtain a trained elastic network regression model.

[0033] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0034] Obtain sample data and an initial elastic network regression model of a sample battery;

[0035] Inputting the sample data into the initial elastic network regression model and outputting a prediction result;

[0036] Determine the target loss of the initial elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term;

[0037] The initial elastic network regression model is trained according to the target loss to obtain a trained elastic network regression model.

[0038] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0039] Obtain sample data and an initial elastic network regression model of a sample battery;

[0040] Inputting the sample data into the initial elastic network regression model and outputting a prediction result;

[0041] Determine the target loss of the initial elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term;

[0042] The initial elastic network regression model is trained according to the target loss to obtain a trained elastic network regression model.

[0043] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0044] Obtain sample data and an initial elastic network regression model of a sample battery;

[0045] Inputting the sample data into the initial elastic network regression model and outputting a prediction result;

[0046] Determine the target loss of the initial elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term;

[0047] The initial elastic network regression model is trained according to the target loss to obtain a trained elastic network regression model.

[0048] The construction method, device and equipment of the elastic network regression model of the battery retention rate mentioned above obtain sample data of the sample battery and the initial elastic network regression model; input the sample data into the initial elastic network regression model and output the prediction result; determine the target loss of the initial elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term; train the initial elastic network regression model according to the target loss to obtain the trained elastic network regression model. By introducing a penalty for the regression coefficient in the regularization term, the complexity of the elastic network regression model is controlled to prevent the elastic network regression model from overfitting, thereby improving the efficiency of training the initial elastic network regression model, and since the number of sample data required for training the regression model is small, the complexity of training the initial elastic network regression model is reduced, thereby further improving the efficiency of training the initial elastic network regression model. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 An application environment diagram of a method for constructing an elastic network regression model of battery retention rate in one embodiment;

[0051] Figure 2 A test result diagram of a trained elastic network regression model in one embodiment;

[0052] Figure 3 A schematic diagram of a process for constructing an elastic network regression model of battery retention rate in one embodiment;

[0053] Figure 4 A schematic flow chart of a method for constructing an elastic network regression model of battery retention rate in another embodiment;

[0054] Figure 5 A schematic flow chart of a method for constructing an elastic network regression model of battery retention rate in another embodiment;

[0055] Figure 6 A schematic flow chart of a method for constructing an elastic network regression model of battery retention rate in another embodiment;

[0056] Figure 7 A schematic flow chart of a method for constructing an elastic network regression model of battery retention rate in another embodiment;

[0057] Figure 8 A schematic flow chart of a method for constructing an elastic network regression model of battery retention rate in another embodiment;

[0058] Fig. 9 A schematic flow chart of a method for constructing an elastic network regression model of battery retention rate in another embodiment;

[0059] Fig.10 A structural block diagram of a device for constructing an elastic network regression model of a battery retention rate in one embodiment;

[0060] Fig.11 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0062] The method for constructing the elastic network regression model of battery retention rate provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. The terminal 102 sends the sample data of the sample battery to the server 104, and the server 104 trains the initial elastic network regression model according to the sample data to obtain the trained elastic network regression model. Among them, the terminal 102 can be but not limited to various acquisition devices, personal computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, projection devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server 104 can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0063] In one embodiment, Figure 2 As shown, a method for constructing an elastic network regression model of battery retention rate is provided, and this method is applied to Figure 1 The following is an example of a server in the example, including:

[0064] S201, obtaining sample data of a sample battery and an initial elastic network regression model.

[0065] Among them, the elastic network regression model is a linear regression model that combines L1 and L2 regularization penalties. It can handle high-dimensional data and features with multicollinearity. L1 regularization is Lasso regression, which focuses on feature selection and produces a sparse model; L2 is ridge regression, which focuses on balanced compression of all features. The elastic network regression model can be flexibly adjusted between Lasso regression and ridge regression.

[0066] In the embodiment of the present application, the sample battery is a lithium-ion battery, and the electrochemical data of the sample battery is obtained, and the electrochemical data of the sample battery is processed to obtain the sample data of the sample battery. The server establishes a linear regression model based on the characteristics of Lasso regression and ridge regression to obtain an initial elastic network regression model. Optionally, the initial elastic network regression model can be shown as Formula 1:

[0067] (Formula 1)

[0068] Among them, y is the target variable, that is, the capacity retention rate, is the intercept term, is the regression coefficient of the jth feature, is the jth feature variable, and p is the number of features.

[0069] Optionally, the server may send a data acquisition request to the terminal, thereby receiving the sample data of the target battery sent by the terminal; or, the server may acquire the electrochemical data of the sample battery from the terminal; or, the user may input the sample data or electrochemical data of the sample battery to the server through the interactive interface. Exemplarily, the terminal may be a third-party device or a data acquisition device, etc.

[0070] Optionally, the electrochemical data of the sample battery can be cleaned to remove irrelevant or redundant features in the electrochemical data of the sample battery, and to process missing values ​​and outliers to obtain processed electrochemical data. Further, the processed electrochemical data can be standardized to eliminate dimensional differences between different features.

[0071] Optionally, feature screening processing may be performed on the electrochemical data or the data after standardization processing to obtain sample data, and the sample data includes effective features that meet preset requirements.

[0072] S202, input the sample data into the initial elastic network regression model and output the prediction result.

[0073] In an embodiment of the present application, the sample data is input into the initial elastic network regression model to obtain the prediction result corresponding to the sample data, that is, the capacity retention rate corresponding to the sample battery is obtained.

[0074] S203, determining the target loss of the initial elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term.

[0075] In the embodiment of the present application, the target loss of the initial elastic network regression model is determined according to the prediction result and the target value of the sample data. Optionally, the target loss can be shown as Formula 2:

[0076] (Formula 2)

[0077] in, is the error term, is the regularization term, and n in the error term is the number of samples. is the target value of the ith sample, is the jth eigenvalue of the i-th sample. The error term is used to measure the mean square error between the model prediction value and the actual value. is the regularization parameter, which is used to control the strength of the regularization term. is a mixing parameter, and , used to adjust the ratio of L1 and L2 regularization. When , the regression type of the initial elastic network regression model is LASSO regression; when When , the regression type of the initial elastic network regression model is ridge regression; when , the regression type of the initial elastic network regression model is elastic network regression.

[0078] S204, training the initial elastic network regression model according to the target loss to obtain a trained elastic network regression model.

[0079] In an embodiment of the present application, the model parameters of the initial elastic network regression model are adjusted according to the target loss to obtain a new initial elastic network regression model; and the step of inputting the effective features of the sample battery into the initial elastic network regression model to obtain the loss function of the elastic network regression model is returned. Optionally, the model parameters may include an intercept term and a regression coefficient of each feature.

[0080] As an optional implementation, it can be determined whether the target loss is less than a preset loss threshold. If so, the current initial elastic network regression model is determined as the trained elastic network regression model; if not, the model parameters of the initial elastic network regression model are adjusted to obtain a new initial elastic network regression model.

[0081] As another optional implementation, the target loss of the current initial elastic network regression model is determined, and the model parameters of the initial elastic network regression model are adjusted to obtain a new initial elastic network regression model until the number of adjustments reaches a preset number of training times, and the elastic network regression model corresponding to the smallest loss among all target losses is determined as the trained elastic network regression model.

[0082] Optionally, part of the sample data can be used as a training set, and the rest of the sample data can be used as test data to test the trained elastic network regression model. The test results are shown in the figure below. Figure 3 As shown in the figure, the root mean square error (RMSE) of the training set is 0.002404, the mean absolute percentage error (MAPE) of the training set is 0.20%, and the coefficient of determination R 2 The RMSE of the test set is 0.003029, the average absolute MAPE of the test set is 0.28%, and the R 2 It is 0.8603.

[0083] In the method for constructing the elastic network regression model of the battery retention rate, sample data of the sample battery and the initial elastic network regression model are obtained; the sample data is input into the initial elastic network regression model, and the prediction result is output; the target loss of the initial elastic network regression model is determined according to the prediction result; the target loss includes an error term and a regularization term; the initial elastic network regression model is trained according to the target loss to obtain the trained elastic network regression model. By introducing a penalty for the regression coefficient in the regularization term, the complexity of the elastic network regression model is controlled to prevent the elastic network regression model from overfitting, thereby improving the efficiency of training the initial elastic network regression model, and since the number of sample data required for training the regression model is small, the complexity of training the initial elastic network regression model is reduced, thereby further improving the efficiency of training the initial elastic network regression model.

[0084] In one embodiment, an implementation of the above S201 is provided, such as Figure 4 As shown, the above-mentioned "obtaining sample data of a sample battery" includes:

[0085] S301, obtaining initial data of a sample battery.

[0086] The initial data includes a plurality of initial features.

[0087] In an embodiment of the present application, the server may send a data acquisition request to the terminal, thereby receiving the initial data of the sample battery sent by the terminal; or, the user may input the initial data of the sample battery to the server through an interactive interface.

[0088] S302, performing correlation screening on a plurality of initial features in the initial data to determine a plurality of intermediate features of the sample battery.

[0089] In an embodiment of the present application, the correlation coefficients between multiple initial features and the target variable are determined, the initial features are sorted from large to small according to the absolute values ​​of the correlation coefficients, and the initial features with correlation coefficients greater than a preset threshold are determined as intermediate features; or, according to the preset number of features, the initial features whose sorting positions are not greater than the preset number of features are determined as intermediate features from the sorting queue of the initial features, that is, the features with higher correlation with the target variable are used as intermediate features.

[0090] Optionally, the correlation coefficient can be determined by determining the covariance between multiple initial features and the target variable; or, the correlation coefficient can be determined by determining the standard deviation between multiple initial features and the target variable; or, the correlation coefficient between multiple initial features and the target variable can be determined according to a preset formula.

[0091] S303, screening the intermediate features according to the multicollinearity of the intermediate features to determine sample data.

[0092] In an embodiment of the present application, the multicollinearity of each intermediate feature is determined, the intermediate features with a high degree of multicollinearity are removed, and the remaining valid intermediate features are determined as sample data.

[0093] Optionally, the pairwise correlation coefficients of the intermediate features may be determined. If the correlation coefficient value between the intermediate features is greater than the correlation coefficient threshold, it is determined that there is a multicollinearity problem between the intermediate features corresponding to the correlation coefficient, and the intermediate features corresponding to the correlation coefficient are removed.

[0094] In the above application embodiment, the initial sample is screened twice for correlation and multicollinearity to determine the sample data, thereby improving the accuracy of the sample data.

[0095] In one embodiment, an implementation of the above S303 is provided, such as Figure 5 As shown, the above “screening the intermediate features according to the multicollinearity of the intermediate features to determine the sample data” includes:

[0096] S401, determining the variance inflation factor of each intermediate feature.

[0097] In the embodiment of the present application, for each intermediate feature , taking the intermediate feature as the dependent variable and other intermediate features as independent variables, a linear regression model is established, so as to determine the variance inflation factor of each intermediate feature according to the linear regression model and the preset variance inflation factor determination formula, wherein the linear regression model can be shown as formula 3:

[0098] (Formula 3)

[0099] S402, determining sample data according to the variance inflation factor of each intermediate feature and the intermediate feature of a first preset threshold.

[0100] In the embodiment of the present application, the intermediate features whose variance inflation factors are less than the first preset threshold are determined as valid features, and the valid features are combined as sample data.

[0101] Optionally, the first preset threshold may be 5 or 10, and the variance inflation factor may be expressed as VIF. The VIF value reflects The larger the VIF value, the greater the degree to which the variance of is inflated due to the linear correlation of other independent variables. The higher the degree of multicollinearity. When VIF = 1, it indicates that there is no multicollinearity between this intermediate feature and other features, indicating that this intermediate feature is completely independent; when 1 < VIF < 5, it indicates that there is a certain degree of multicollinearity between this intermediate feature and other features; when VIF ≥ 5 or VIF ≥ 10, it indicates that there is strong multicollinearity between this intermediate feature and other features.

[0102] In the above application embodiments, the variance inflation factor of each intermediate feature is determined, so as to remove the intermediate feature regions with variance inflation factors greater than the first preset threshold, and obtain sample data, improving the reliability and accuracy of the sample data.

[0103] In one embodiment, an implementation manner of the above S401 is provided, as Figure 6 shown, the above "determining the variance inflation factor of each intermediate feature" includes:

[0104] S501, according to the average value of each intermediate feature and each intermediate feature, determine the data fluctuation amount and data variation amount of each intermediate feature.

[0105] In the embodiments of the present application, for each intermediate feature, for example, the jth intermediate feature, the process of determining the data fluctuation amount can be as shown in Equation 4, and the process of determining the data variation amount can be as shown in Equation 5:

[0106] (Equation 4)

[0107] (Equation 5)

[0108] Among them, SST is the data fluctuation amount of this intermediate feature, used to represent the total data fluctuation amount, SSR is the data variation amount, is the average value of multiple feature values of this intermediate feature, is the model prediction value of the jth intermediate feature, is the ith feature value of this intermediate feature.

[0109] S502, according to the data fluctuation amount and data variation amount of each intermediate feature, determine the determination coefficient of each intermediate feature.

[0110] In the embodiments of the present application, according to the ratio of the data fluctuation amount and the data variation amount, determine the determination coefficient R 2 of each intermediate feature. Exemplarily, the process of determining the determination coefficient is as shown in Equation 6:

[0111] (Equation 6)

[0112] It should be noted that the coefficient of determination is an indicator of the goodness of fit of the model, which is used to evaluate the degree to which the model explains the variation of the actual values. The value of R² is between [0,1]. If R²=1, it means that the model can fully explain all the variations of the target variable; if R²=0, it means that the model has no explanatory power.

[0113] S503: Determine the variance inflation factor of each intermediate feature according to the square value of the coefficient of determination of each intermediate feature.

[0114] In the embodiment of the present application, the larger the square value of the determination coefficient, the smaller the variance inflation factor; the smaller the square value of the determination coefficient, the larger the variance inflation factor. Exemplarily, the determination process of the variance inflation factor of the jth intermediate feature is shown in Formula 7:

[0115] (Formula 7)

[0116] In the above application embodiment, the determination coefficient corresponding to the intermediate feature is first determined, and then the variance inflation factor of the intermediate feature is determined based on the determination coefficient, thereby improving the accuracy of the variance inflation factor. Compared with directly using the determination coefficient to determine the multicollinearity of the intermediate feature, determining the multicollinearity of the intermediate feature through the variance inflation factor is more accurate.

[0117] In one embodiment, an implementation of the above S302 is provided, such as Figure 7 As shown, the above-mentioned "screening the correlation of multiple initial features in the initial data to determine multiple intermediate features of the sample battery" includes:

[0118] S601, determining the Pearson correlation coefficient between each initial characteristic and the capacity retention rate.

[0119] In the embodiment of the present application, the Pearson correlation coefficient between the multiple initial features and the target variable is determined. Wherein the target variable is the capacity retention rate, and illustratively, the determination process of the Pearson correlation coefficient can be shown as Formula 8:

[0120] (Formula 8)

[0121] in, is the Pearson correlation coefficient between the jth feature and the target variable, is the j-th eigenvalue of the ith sample, is the target variable value of the i-th sample, and are the means of the feature and target variables, respectively.

[0122] S602: Determine the initial feature whose Pearson correlation coefficient is greater than the second preset threshold as an intermediate feature.

[0123] In the embodiment of the present application, each Pearson correlation coefficient is compared with the second preset threshold, so that the initial feature whose Pearson correlation coefficient is greater than the second preset threshold is determined as the intermediate feature.

[0124] In the above application embodiment, the Pearson correlation coefficient between each initial feature and the preset capacity retention rate is determined, and then the correlation between each initial feature and the capacity retention rate is determined according to each Pearson correlation coefficient, and the feature with a higher correlation with the capacity retention rate is used as the intermediate feature, thereby improving the accuracy of the intermediate feature.

[0125] In one embodiment, Figure 8 As shown, the method for constructing the elastic network regression model of the battery retention rate also includes:

[0126] S205, acquiring characteristic information of the battery to be predicted, and performing standardization processing on the characteristic information of the battery to be predicted to obtain processed characteristic information.

[0127] In the embodiment of the present application, each acquired feature information is standardized. For example, the j-th feature information is standardized as shown in Formula 9:

[0128] (Formula 9)

[0129] in, is the jth feature information after standardization, is the jth feature information of the battery to be predicted, is the average value of the j-th feature information determined during the training of the initial elastic network regression model, is the standard deviation of the j-th feature information determined during the training of the initial elastic net regression model.

[0130] S206, inputting the processed feature information into the trained elastic network regression model to determine the capacity retention rate of the battery to be predicted.

[0131] In an embodiment of the present application, as shown in Formula 10, the processed feature information is input into the trained elastic network regression model, and the elastic network regression model analyzes and predicts the processed feature information to determine the capacity retention rate of the battery to be predicted.

[0132] (Formula 10)

[0133] in, is the capacity retention rate of the battery to be predicted.

[0134] In the above application embodiment, characteristic information can be determined based on data of any charge and discharge cycle of the battery to be tested, without collecting a large amount of charge and discharge data of the battery to be tested for prediction, thereby improving the efficiency of predicting the capacity retention rate of the battery to be predicted.

[0135] In one embodiment, a method for constructing a complete elastic network regression model of battery retention rate is provided, such as Fig. 9 As shown, the above method includes:

[0136] S1, obtain the initial data of the sample battery and the initial elastic network regression model.

[0137] S2, determine the Pearson correlation coefficient between each initial characteristic and the capacity retention rate.

[0138] S3, determining the initial feature whose Pearson correlation coefficient is greater than the second preset threshold as the intermediate feature.

[0139] S4, determining the data fluctuation amount and data variation amount of each intermediate feature according to the average value of each intermediate feature and each intermediate feature.

[0140] S5, determining the coefficient of determination of each intermediate feature according to the data fluctuation amount and data variation amount of each intermediate feature.

[0141] S6, determining the variance inflation factor of each intermediate feature according to the square value of the determination coefficient of each intermediate feature.

[0142] S7, determining sample data according to the variance inflation factor of each intermediate feature and the intermediate feature of the first preset threshold.

[0143] S8, input the sample data into the initial elastic network regression model and output the prediction results.

[0144] S9, determining the target loss of the initial elastic network regression model according to the prediction results; the target loss includes an error term and a regularization term.

[0145] S10, training the initial elastic network regression model according to the target loss to obtain a trained elastic network regression model.

[0146] S11, acquiring characteristic information of the battery to be predicted, and performing standardization processing on the characteristic information of the battery to be predicted to obtain processed characteristic information.

[0147] S12, inputting the processed feature information into the trained elastic network regression model to determine the capacity retention rate of the battery to be predicted.

[0148] In the method for constructing the elastic network regression model of the battery retention rate, sample data of the sample battery and the initial elastic network regression model are obtained; the sample data is input into the initial elastic network regression model, and the prediction result is output; the target loss of the initial elastic network regression model is determined according to the prediction result; the target loss includes an error term and a regularization term; the initial elastic network regression model is trained according to the target loss to obtain the trained elastic network regression model. By introducing a penalty for the regression coefficient in the regularization term, the complexity of the elastic network regression model is controlled to prevent the elastic network regression model from overfitting, thereby improving the efficiency of training the initial elastic network regression model, and since the number of sample data required for training the regression model is small, the complexity of training the initial elastic network regression model is reduced, thereby further improving the efficiency of training the initial elastic network regression model.

[0149] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0150] Based on the same inventive concept, the embodiment of the present application also provides a device for constructing an elastic network regression model of battery retention rate for implementing the method for constructing an elastic network regression model of battery retention rate involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of the device for constructing one or more elastic network regression models of battery retention rate provided below can be referred to the limitations of the method for constructing an elastic network regression model of battery retention rate above, and will not be repeated here.

[0151] In one embodiment, Fig.10 As shown, a device for constructing an elastic network regression model of battery retention rate is provided, comprising: an acquisition module 10, a first prediction module 11, a determination module 12 and a training module 13, wherein:

[0152] The acquisition module 10 is used to acquire sample data of the sample battery and an initial elastic network regression model.

[0153] The first prediction module 11 is used to input sample data into the initial elastic network regression model and output prediction results.

[0154] The determination module 12 is used to determine the target loss of the elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term.

[0155] The training module 13 is used to train the initial elastic network regression model according to the target loss to obtain a trained elastic network regression model.

[0156] In one embodiment, the acquisition module 10 includes: an acquisition unit, a first screening unit and a second screening unit, wherein:

[0157] The acquisition unit is used to acquire initial data of the sample battery.

[0158] The first screening unit is used to perform correlation screening on multiple initial features in the initial data to determine multiple intermediate features of the sample battery.

[0159] The second screening unit is used to screen the intermediate features according to the multicollinearity of the intermediate features to determine the sample data.

[0160] In one embodiment, the second screening unit is specifically used to determine the variance inflation factor of each intermediate feature; and determine the sample data according to the variance inflation factor of each intermediate feature and the intermediate feature of the first preset threshold.

[0161] In one embodiment, the above-mentioned second screening unit is specifically used to determine the data fluctuation and data variation of each intermediate feature based on the average value of each intermediate feature and each intermediate feature; determine the determination coefficient of each intermediate feature based on the data fluctuation and data variation of each intermediate feature; determine the variance inflation factor of each intermediate feature based on the square value of the determination coefficient of each intermediate feature.

[0162] In one embodiment, the first screening unit is specifically used to determine the Pearson correlation coefficient between each initial feature and the capacity retention rate; and the initial feature whose Pearson correlation coefficient is greater than the second preset threshold is determined as the intermediate feature.

[0163] In one embodiment, the apparatus for constructing the elastic network regression model of the battery retention rate further includes: a processing module and a second prediction module, wherein:

[0164] The processing module is used to obtain the characteristic information of the battery to be predicted, and perform standardization processing on the characteristic information of the battery to be predicted to obtain the processed characteristic information.

[0165] The second prediction module is used to input the processed feature information into the trained elastic network regression model to determine the capacity retention rate of the battery to be predicted.

[0166] Each module in the construction device of the elastic network regression model of the battery retention rate can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0167] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Fig.11 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the construction data of the elastic network regression model of the battery retention rate. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for constructing an elastic network regression model of the battery retention rate is implemented.

[0168] Those skilled in the art will understand that Fig.11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0169] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0170] Obtain sample data and an initial elastic network regression model of a sample battery;

[0171] Input sample data into the initial elastic network regression model and output the prediction results;

[0172] Determine the target loss of the initial elastic network regression model according to the prediction results; the target loss includes an error term and a regularization term;

[0173] The initial elastic network regression model is trained according to the target loss to obtain a trained elastic network regression model.

[0174] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0175] Obtaining initial data of sample batteries;

[0176] Perform correlation screening on multiple initial features in the initial data to determine multiple intermediate features of the sample battery;

[0177] The intermediate features are screened according to their multicollinearity to determine the sample data.

[0178] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0179] Determine the variance inflation factor for each intermediate feature;

[0180] Sample data is determined according to the variance inflation factor of each intermediate feature and the intermediate feature of a first preset threshold.

[0181] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0182] According to the average value of each intermediate feature and each intermediate feature, the data fluctuation amount and data variation amount of each intermediate feature are determined;

[0183] According to the data fluctuation and data variation of each intermediate feature, the determination coefficient of each intermediate feature is determined;

[0184] The variance inflation factor of each intermediate feature is determined according to the square value of the determination coefficient of each intermediate feature.

[0185] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0186] Determine the Pearson correlation coefficient between each initial characteristic and capacity retention;

[0187] The initial features whose Pearson correlation coefficient is greater than the second preset threshold are determined as intermediate features.

[0188] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0189] Acquire characteristic information of the battery to be predicted, and perform standardization processing on the characteristic information of the battery to be predicted to obtain processed characteristic information;

[0190] The processed feature information is input into the trained elastic network regression model to determine the capacity retention rate of the battery to be predicted.

[0191] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0192] Obtain sample data and an initial elastic network regression model of a sample battery;

[0193] Input sample data into the initial elastic network regression model and output the prediction results;

[0194] Determine the target loss of the initial elastic network regression model according to the prediction results; the target loss includes an error term and a regularization term;

[0195] The initial elastic network regression model is trained according to the target loss to obtain a trained elastic network regression model.

[0196] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0197] Obtaining initial data of sample batteries;

[0198] Perform correlation screening on multiple initial features in the initial data to determine multiple intermediate features of the sample battery;

[0199] The intermediate features are screened according to their multicollinearity to determine the sample data.

[0200] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0201] Determine the variance inflation factor for each intermediate feature;

[0202] Sample data is determined according to the variance inflation factor of each intermediate feature and the intermediate feature of a first preset threshold.

[0203] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0204] According to the average value of each intermediate feature and each intermediate feature, the data fluctuation amount and data variation amount of each intermediate feature are determined;

[0205] According to the data fluctuation and data variation of each intermediate feature, the determination coefficient of each intermediate feature is determined;

[0206] The variance inflation factor of each intermediate feature is determined according to the square value of the determination coefficient of each intermediate feature.

[0207] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0208] Determine the Pearson correlation coefficient between each initial characteristic and capacity retention;

[0209] The initial features whose Pearson correlation coefficient is greater than the second preset threshold are determined as intermediate features.

[0210] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0211] Acquire characteristic information of the battery to be predicted, and perform standardization processing on the characteristic information of the battery to be predicted to obtain processed characteristic information;

[0212] The processed feature information is input into the trained elastic network regression model to determine the capacity retention rate of the battery to be predicted.

[0213] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0214] Obtain sample data and an initial elastic network regression model of a sample battery;

[0215] Input sample data into the initial elastic network regression model and output the prediction results;

[0216] Determine the target loss of the initial elastic network regression model according to the prediction results; the target loss includes an error term and a regularization term;

[0217] The initial elastic network regression model is trained according to the target loss to obtain a trained elastic network regression model.

[0218] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0219] Obtaining initial data of sample batteries;

[0220] Perform correlation screening on multiple initial features in the initial data to determine multiple intermediate features of the sample battery;

[0221] The intermediate features are screened according to their multicollinearity to determine the sample data.

[0222] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0223] Determine the variance inflation factor for each intermediate feature;

[0224] Sample data is determined according to the variance inflation factor of each intermediate feature and the intermediate feature of a first preset threshold.

[0225] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0226] According to the average value of each intermediate feature and each intermediate feature, the data fluctuation amount and data variation amount of each intermediate feature are determined;

[0227] According to the data fluctuation and data variation of each intermediate feature, the determination coefficient of each intermediate feature is determined;

[0228] The variance inflation factor of each intermediate feature is determined according to the square value of the determination coefficient of each intermediate feature.

[0229] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0230] Determine the Pearson correlation coefficient between each initial characteristic and capacity retention;

[0231] The initial features whose Pearson correlation coefficient is greater than the second preset threshold are determined as intermediate features.

[0232] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0233] Acquire characteristic information of the battery to be predicted, and perform standardization processing on the characteristic information of the battery to be predicted to obtain processed characteristic information;

[0234] The processed feature information is input into the trained elastic network regression model to determine the capacity retention rate of the battery to be predicted.

[0235] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0236] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0237] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for constructing an elastic network regression model of battery retention rate, characterized in that: The method comprises: Obtain sample data and an initial elastic network regression model of a sample battery; Inputting the sample data into the initial elastic network regression model and outputting a prediction result; Determine the target loss of the initial elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term; The initial elastic network regression model is trained according to the target loss to obtain a trained elastic network regression model.

2. The method according to claim 1, characterized in that The obtaining of sample data of the sample battery includes: Acquiring initial data of the sample battery; Performing correlation screening on a plurality of initial features in the initial data to determine a plurality of intermediate features of the sample battery; The intermediate features are screened according to the multicollinearity of the intermediate features to determine the sample data.

3. The method according to claim 2, characterized in that The step of screening the intermediate features according to the multicollinearity of the intermediate features to determine the sample data includes: Determining a variance inflation factor for each of the intermediate features; The sample data is determined according to the variance inflation factor of each of the intermediate features and the intermediate features of a first preset threshold.

4. The method according to claim 3, characterized in that Determining the variance inflation factor of each of the intermediate features includes: Determine the data fluctuation amount and data variation amount of each intermediate feature according to the average value of each intermediate feature and each intermediate feature; Determining the coefficient of determination of each intermediate feature according to the data fluctuation amount and the data variation amount of each intermediate feature; The variance inflation factor of each intermediate feature is determined according to the square value of the determination coefficient of each intermediate feature.

5. The method according to claim 2, characterized in that: The step of performing correlation screening on the multiple initial features in the initial data to determine the multiple intermediate features of the sample battery includes: Determining the Pearson correlation coefficient between each of the initial characteristics and the capacity retention rate; The initial feature whose Pearson correlation coefficient is greater than a second preset threshold is determined as the intermediate feature.

6. The method according to any one of claims 1 to 5, characterized in that: The method further comprises: Acquiring characteristic information of a battery to be predicted, and performing standardization processing on the characteristic information of the battery to be predicted to obtain processed characteristic information; The processed feature information is input into the trained elastic network regression model to determine the capacity retention rate of the battery to be predicted.

7. A device for constructing an elastic network regression model of battery retention rate, characterized in that: The device comprises: An acquisition module, used to acquire sample data of a sample battery and an initial elastic network regression model; A first prediction module, used for inputting the sample data into the initial elastic network regression model and outputting a prediction result; A determination module, used to determine the target loss of the elastic network regression model according to the prediction result; the target loss includes an error term and a regularization term; A training module is used to train the initial elastic network regression model according to the target loss to obtain a trained elastic network regression model.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images