Method for predicting open-circuit voltage of battery

Abstract

The invention discloses a battery open-circuit voltage prediction method, and relates to the technical field of battery open-circuit voltage prediction.The battery open-circuit voltage prediction method comprises battery data acquisition, battery state analysis, prediction method distribution, open-circuit voltage prediction and terminal display, by acquiring rated data and real-time data of a battery, the basic performance and the current state of the battery are comprehensively known, and the battery open-circuit voltage prediction accuracy is improved. The open-circuit voltage can be predicted more accurately; the voltage change and other key parameters of the battery are obtained in real time, the running state and health degree of the battery can be accurately analyzed, and potential problems of the battery can be found in time; according to the operation state evaluation coefficient and the final voltage of the battery, the most suitable open-circuit voltage prediction method is flexibly selected, the prediction accuracy is improved, the high efficiency of the prediction process is ensured, and computing resources are effectively utilized; a powerful guarantee is provided for efficient operation of battery open-circuit voltage prediction, and the voltage prediction level is improved.

Description

Technical Field

[0001] The present application relates to the technical field of battery open circuit voltage prediction, and in particular to a method for battery open circuit voltage prediction. Background Art

[0002] Open circuit voltage refers to the potential difference between the positive and negative electrodes of a battery when no current flows through it. It reflects the electrochemical state inside the battery and is one of the important parameters of battery performance. Therefore, it is very important to predict open circuit voltage.

[0003] Prior art, such as the invention application patent with announcement number: CN118011233A, discloses a method, system and device for predicting the open circuit voltage of a battery, which relates to the field of battery technology. The method includes: obtaining a sample data set of standby batteries, each battery sample includes a standby data subset and a corresponding OCV value; the standby data subset includes a voltage value, a current value, a battery temperature, a SOC value and a SOH value; using each standby data subset in the standby battery sample data set as input, and the corresponding OCV value as a label, to train a preset network to obtain an open circuit voltage prediction model; deploying the open circuit voltage prediction model on a control platform; obtaining the current standby data subset of the target battery through the control platform, and inputting the current standby data subset into the open circuit voltage prediction model to obtain the current OCV value of the target battery. The present invention can determine the accurate open circuit voltage of the battery in real time online.

[0004] In view of the above scheme, the inventors of the present application have found that the above technology has at least the following technical problems: 1. The current prediction and analysis of the battery open circuit voltage is mainly carried out through SOC analysis, but in the process of analyzing SOC, the prediction method is not selected according to the state of the battery, and there is no basic prediction method, data-driven prediction method and real-time prediction method, so it is impossible to more accurately understand the actual situation of the current battery SOC, and it is also impossible to ensure the comprehensiveness and authenticity of the battery open circuit voltage analysis process, and thus it is impossible to ensure the reference and accuracy of the analysis results, and it is impossible to provide a reliable basis for the optimization of the analysis process.

[0005] 2. Currently, there is a lack of calculation of the battery health status SOH based on the change of internal resistance, which cannot truly display the current situation of the battery rated capacity, and thus cannot provide important reference for battery management, cannot compensate for the impact of current fluctuations on SOC estimation, and cannot provide stable and reliable SOC predictions on a longer time scale. It cannot solve the limitations of the current battery open circuit voltage prediction process and cannot guarantee the reliability and authenticity of the battery open circuit voltage prediction analysis results. Summary of the invention

[0006] In view of the above-mentioned technical deficiencies, the purpose of this application is to provide a method for predicting the open circuit voltage of a battery.

[0007] To solve the above technical problems, the present application adopts the following technical solutions: The present application provides a method for predicting the open circuit voltage of a battery, the method comprising the following steps: Step 1, battery data acquisition: When predicting the open circuit voltage of each battery, obtain the rated data of each battery, and obtain the initial voltage and final voltage of each battery in the current time period.

[0008] Step 2: Battery status analysis: Get the real-time data of each battery, analyze and derive the operating status evaluation coefficient and health status evaluation coefficient of each battery according to the initial voltage and final voltage of each battery in the current time period, and determine the current operating status of each battery.

[0009] Step 3: Allocation of prediction methods: According to the operating status evaluation coefficient and final voltage of each battery, the open circuit voltage prediction method of each battery is allocated to obtain the open circuit voltage prediction method of each battery.

[0010] Step 4: Open circuit voltage prediction: According to the open circuit voltage prediction method of each battery, the functional relationship between the state of charge and the open circuit voltage, and the relationship model between internal resistance, temperature and state of charge, the open circuit voltage of each battery is predicted.

[0011] Step 5: Display terminal: Display the open circuit voltage of each battery in the form of a graph.

[0012] Preferably, the rated data of the battery includes the initial state of charge, rated capacity, rated internal resistance and rated number of cycles of the battery.

[0013] Preferably, the real-time data of the battery includes the current remaining available power, the current internal resistance value, the current temperature, the current value, the working time, the number of deep cycles and the number of shallow cycles.

[0014] Preferably, the open circuit voltage prediction method includes a basic prediction method, a data-driven prediction method and a real-time prediction method.

[0015] The beneficial effects of the present application are as follows: 1. The present application provides a method for predicting the open circuit voltage of a battery, which obtains the rated data and real-time data of the battery to fully understand the basic performance and current status of the battery, and thus more accurately predicts its open circuit voltage; real-time acquisition of the voltage change and other key parameters of the battery can accurately analyze the operating status and health of the battery, and help to promptly discover potential problems with the battery; according to the battery operating status evaluation coefficient and final voltage, the most suitable open circuit voltage prediction method can be flexibly selected, which not only improves the accuracy of the prediction, but also ensures the efficiency of the prediction process and effectively utilizes computing resources; it provides a strong guarantee for the efficient operation of the battery open circuit voltage prediction and improves the level of voltage prediction.

[0016] 2. Based on real-time voltage data combined with initial voltage and final voltage, this application can quickly and accurately determine whether the battery is currently in charging, discharging or static state, which is crucial to the working mode of the battery; further, combined with the rated parameters of the battery and real-time acquired dynamic data such as current and temperature, the internal resistance is used for comprehensive analysis to accurately evaluate the health status of the battery, providing a scientific basis for prediction.

[0017] 3. In this application, the most appropriate open circuit voltage prediction method is flexibly selected according to the actual operating status of the battery, which reflects a high degree of intelligence and adaptability. The basic prediction method is suitable for when the battery is at rest or in a low current state, and can accurately reflect the charge state of the battery; when the battery is in working condition, the data-driven prediction method or the real-time prediction method is intelligently selected according to the final voltage to overcome the influence of current fluctuations on charge state estimation and ensure the accuracy and reliability of the prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 The figure is a flowchart of the implementation steps of the present application method. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0021] See also Figure 1 As shown, the present application provides a method for predicting the open circuit voltage of a battery, comprising the following steps: Step 1, battery data acquisition: when predicting the open circuit voltage of each battery, obtain the rated data of each battery, and obtain the initial voltage and final voltage of each battery in the current time period.

[0022] In a specific example, the rated data of the battery includes the initial state of charge, rated capacity, rated internal resistance and rated number of cycles of the battery.

[0023] Step 2: Battery status analysis: Get the real-time data of each battery, analyze and derive the operating status evaluation coefficient and health status evaluation coefficient of each battery according to the initial voltage and final voltage of each battery in the current time period, and determine the current operating status of each battery.

[0024] In a specific example, the real-time data of the battery includes the current remaining available power, the current internal resistance value, the current temperature, the current value, the working time, the number of deep cycles and the number of shallow cycles.

[0025] In a specific example, the analysis obtains the operating status evaluation coefficient of each battery. The specific analysis process is as follows: the initial voltage and the final voltage of each battery are respectively recorded as CY i and ZY i , where i=1,2,......,n, i represents the serial number corresponding to each battery, n represents the total number of batteries, and n represents a natural integer greater than or equal to 2.

[0026] According to the calculation formula Obtain the operating status evaluation coefficient λ of each battery i , where Y′ represents the base value of voltage, σ i It is expressed as the correction factor corresponding to the voltage of each battery.

[0027] It should be noted that the value of Y′ is 1 volt, which is used to eliminate the unit of voltage in the calculation process.

[0028] It should be noted that the correction factor corresponding to the voltage of each battery is a digital factor multiplied by the uncorrected measurement result in order to correct the error of the voltage of each battery; it is obtained by analyzing the measured value and actual value of the voltage of each battery in the historical records, wherein according to the calculation formula:

[0029] The correction factor σ corresponding to the voltage of each battery is obtained by analysis. i .

[0030] In a specific example, the analysis obtains the health status assessment coefficient of each battery, and the specific analysis process is as follows: the rated capacity, rated internal resistance, rated number of cycles, current remaining available power, current internal resistance value, deep cycle number and shallow cycle number of each battery are recorded as ER i ,EN i 、EX i ,DR i DN i SX i and QX i .

[0031] According to the calculation formula Obtain the health status assessment coefficient β of each batteryi , where ω1 and ω2 represent the weight factors corresponding to the number of deep cycles and the number of shallow cycles, respectively.

[0032] It should be noted that 0<ω1<1, 0<ω2<1, ω1+ω2=1.

[0033] It should be noted that the weight factors corresponding to the number of deep cycles and the weight factors corresponding to the number of shallow cycles are obtained through factor analysis. First, the information of the number of deep cycles and the number of shallow cycles is condensed, and then the variance explanation rate after rotation is obtained, and the weight is obtained by dividing the cumulative variance explanation rate.

[0034] It should be noted that factor analysis is a well-known technology. It is a multivariate statistical analysis method that starts from studying the internal dependencies of variables and reduces some variables with intricate relationships to a few comprehensive factors; information concentration is expressed as calculating the median; the variance explanation rate is the amount of information extracted by the factor, and the variance explanation rate = characteristic root / total number of analysis items; the variance explanation rate after rotation is expressed as the variance explanation rate of the factor after maximum variance rotation.

[0035] In a specific example, the specific judgment process for judging the current operating status of each battery is as follows: comparing the operating status evaluation coefficient of each battery with the threshold of the operating status evaluation coefficient in the database; when the operating status evaluation coefficient of a battery is equal to the threshold of the operating status evaluation coefficient in the database, the operating status of the battery is judged to be stationary; when the operating status evaluation coefficient of a battery is less than the threshold of the operating status evaluation coefficient in the database, the operating status of the battery is judged to be charging; when the operating status evaluation coefficient of a battery is greater than the threshold of the operating status evaluation coefficient in the database, the operating status of the battery is judged to be discharging.

[0036] It should be noted that the threshold value of the operating state evaluation coefficient is expressed as a value corresponding to when the operating state of the battery is static, and is infinitely close to 0.

[0037] Based on real-time voltage data combined with initial voltage and final voltage, this application can quickly and accurately determine whether the battery is currently in charging, discharging or static state, which is crucial to the working mode of the battery; further, by combining the rated parameters of the battery with real-time acquired dynamic data such as current and temperature, and using comprehensive analysis of internal resistance, the health status value of the battery can be accurately evaluated, providing a scientific basis for prediction.

[0038] Step 3: Allocation of prediction methods: According to the operating status evaluation coefficient and final voltage of each battery, the open circuit voltage prediction method of each battery is allocated to obtain the open circuit voltage prediction method of each battery.

[0039] In a specific example, the open circuit voltage prediction method includes a basic prediction method, a data-driven prediction method and a real-time prediction method.

[0040] In a specific example, the open circuit voltage prediction method of each battery is allocated, and the specific allocation process is as follows: when the operating state of a battery is static, the basic prediction method is used as the open circuit voltage prediction method of the battery; when the operating state of a battery is charging or discharging, the final voltage of the battery is compared with the predicted voltage threshold in the database, when the final voltage of the battery is less than or equal to the predicted voltage threshold in the database, the data-driven prediction method is used to analyze and obtain the battery voltage, and when the final voltage of the battery is greater than the predicted voltage threshold in the database, the real-time prediction method is used to analyze and obtain the battery voltage.

[0041] In this application, the most appropriate open circuit voltage prediction method is flexibly selected according to the actual operating status of the battery, which reflects a high degree of intelligence and adaptability. The basic prediction method is suitable for when the battery is at rest or in a low current state, and can accurately reflect the charge state of the battery; when the battery is in a working state, the data-driven prediction method or the real-time prediction method is intelligently selected according to the final voltage to overcome the influence of current fluctuations on the charge state estimation, thereby ensuring the accuracy and reliability of the prediction results.

[0042] Step 4: Open circuit voltage prediction: According to the open circuit voltage prediction method of each battery, the functional relationship between the state of charge and the open circuit voltage, and the relationship model between internal resistance, temperature and state of charge, the open circuit voltage of each battery is predicted.

[0043] It should be noted that the process of obtaining the functional relationship between the state of charge and the open circuit voltage of the battery is as follows: use a battery charge and discharge tester, a voltage sensor, a temperature sensor and other equipment as data acquisition equipment to collect data; first, perform standard charging on the battery to ensure that the battery is fully charged to a state of charge of 100%, let the battery stand for 3 hours, and measure the open circuit voltage at this time point; start the discharge test, control the discharge current, and discharge the battery intermittently. After each discharge, let the battery stand and measure the open circuit voltage. Repeat this process ten times to obtain different state of charge values; at the same time, calculate the state of charge value after each discharge based on the discharge current and time, and record the open circuit voltage after each discharge; perform data fitting based on the processed data. First, based on the experimental data, assume that there is a certain functional relationship between the state of charge and the open circuit voltage, such as a polynomial function, and then use MATLAB mathematical software to obtain the polynomial coefficients of the fitting curve through the data fitting command polyfit; According to the coefficients obtained by fitting, construct a functional relationship between the state of charge and the open circuit voltage, such as OCV=a*SOC n +b*SOC n-1+......+d, where n is the degree of the polynomial, and a, b, ..., d are the coefficients obtained by fitting; finally, verification and optimization are carried out, and the accuracy of the fitting result is verified by comparing the degree of fit between the experimental data and the fitting curve. If the fitting accuracy does not meet the requirements, the degree of the polynomial is adjusted or other fitting methods are used to improve the fitting accuracy.

[0044] In a specific example, the open circuit voltage of each battery is predicted according to the open circuit voltage prediction method of each battery. The specific prediction process is as follows: A1. Basic prediction method, according to the calculation formula Get the SOC of each battery i , where ρ i It is expressed as a correction factor for the influence of the health status assessment coefficient of each battery on the rated capacity of the battery, and according to the state of charge of each battery, in the functional relationship between the state of charge of each battery and the open circuit voltage, the open circuit voltage value corresponding to the state of charge of each battery is obtained, and the open circuit voltage value is used as the battery open circuit voltage of each battery.

[0045] A2. Data-driven prediction method first transmits the current internal resistance value and current temperature of each battery to the relationship model of internal resistance, temperature and state of charge, and then outputs the state of charge of each battery, so as to obtain the open circuit voltage value corresponding to the state of charge of each battery in the functional relationship between the state of charge of each battery and the open circuit voltage, and use the open circuit voltage value as the battery open circuit voltage of each battery.

[0046] A3. Real-time prediction method, according to the calculation formula: The battery state of charge is analyzed, where CSOC i Represents the initial state of charge of each battery, CSOC i Expressed as a percentage of the rated capacity of each battery, I i Expressed as the current value of each battery, t i It is expressed as the working time of each battery, so that according to the charge state of each battery, in the functional relationship between the charge state of each battery and the open circuit voltage, the open circuit voltage value corresponding to the charge state of each battery is obtained, and the open circuit voltage value is used as the battery open circuit voltage of each battery.

[0047] It should be noted that the percentage of battery rated capacity is calculated according to the formula:

[0048] ERB i =1*10 -13 *(SX i +QX i ) 4 -8*10 -8 *(SX i +QX i ) 3 +2*10-5 *(SX i +QX i ) 2 -1.68*10 -2 *(SX i +QX i )+99.865 is analyzed.

[0049] It should be noted that the initial voltage and final voltage of the battery in the current time period, the initial voltage of the battery is represented by the battery voltage at the start time point of the current time period, and the final voltage of the battery is represented by the battery voltage at the end time point of the current time period, which are obtained in battery management.

[0050] It should be noted that the rated data of the battery is the factory data inherent to the battery.

[0051] It should be noted that the state of charge is expressed as the state of charge, and the open circuit voltage is expressed as the open circuit voltage.

[0052] In a specific example, the relationship model between temperature and state of charge is specifically acquired through the following process: using a temperature sensor and an internal resistance meter as data acquisition devices; randomly selecting temperature and state of charge from the temperature set and the state of charge set as a set of experimental conditions, thereby obtaining different sets of experimental conditions, and conducting experiments on the battery according to each set of experimental conditions, and then recording the internal resistance value of the battery under each set of experimental conditions during the experiment, and taking the temperature, state of charge and internal resistance value as a data set, thereby obtaining each data set; performing a data check, deleting data sets containing abnormal values ​​or erroneous data, and then using a neural network of a nonlinear regression model to establish a model; first dividing the data set, dividing each data set into a training set and a test set, using the training set data to train the selected model, and during the training process, adjusting the parameters of the model to optimize the performance, and using the test set data to verify the performance of the model, and optimizing the parameters of the model through a cross-validation method, thereby obtaining a relationship model between internal resistance, temperature and state of charge.

[0053] It should be noted that, because there may be a complex nonlinear relationship between internal resistance and temperature and state of charge, a neural network of a nonlinear regression model is used to establish the model.

[0054] Step 5: Display terminal: Display the open circuit voltage of each battery in the form of a graph.

[0055] The present application provides a method for predicting the open circuit voltage of a battery. By acquiring the rated data and real-time data of the battery, the basic performance and current status of the battery can be fully understood, thereby more accurately predicting its open circuit voltage. The voltage change and other key parameters of the battery can be acquired in real time, so that the operating status and health of the battery can be accurately analyzed, which helps to timely discover potential problems of the battery. According to the operating status evaluation coefficient and final voltage of the battery, the most suitable open circuit voltage prediction method can be flexibly selected, which not only improves the accuracy of the prediction, but also ensures the efficiency of the prediction process and effectively utilizes computing resources. This provides a strong guarantee for the efficient operation of the battery open circuit voltage prediction and improves the level of voltage prediction.

[0056] The above contents are merely examples and explanations of the concept of the present application. The technical personnel in the relevant technical field may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the scope of protection of the present application.

Claims

1. A method for predicting battery open circuit voltage, characterized in that: include: Step 1: Obtaining battery data: When predicting the open circuit voltage of each battery, obtain the rated data of each battery, and obtain the initial voltage and final voltage of each battery in the current time period; Step 2: Battery status analysis: obtain real-time data of each battery, analyze and obtain the operating status evaluation coefficient and health status evaluation coefficient of each battery according to the initial voltage and final voltage of each battery in the current time period, and determine the current operating status of each battery; Step 3: Assigning prediction methods: Assigning the open circuit voltage prediction methods of each battery according to the operating status evaluation coefficient and the final voltage of each battery, thereby obtaining the open circuit voltage prediction methods of each battery; Step 4: Open circuit voltage prediction: predict the open circuit voltage of each battery according to the open circuit voltage prediction method of each battery, the functional relationship between the state of charge and the open circuit voltage, and the relationship model between internal resistance, temperature and the state of charge; Step 5: Display terminal: Display the open circuit voltage of each battery in the form of a graph.

2. A method for predicting battery open circuit voltage according to claim 1, characterized in that: The rated data of the battery include the initial state of charge, rated capacity, rated internal resistance and rated number of cycles of the battery.

3. A method for predicting a battery open circuit voltage according to claim 2, characterized in that: The real-time data of the battery includes the current remaining available power, the current internal resistance value, the current temperature, the current value, the working time, the number of deep cycles and the number of shallow cycles.

4. A method for predicting battery open circuit voltage according to claim 3, characterized in that: The analysis results in the operating status evaluation coefficient of each battery. The specific analysis process is as follows: The initial voltage and final voltage of each battery are denoted as CY i and ZY i , where i = 1, 2, ..., n, i represents the number corresponding to each battery, n represents the total number of batteries, and n represents a natural integer greater than or equal to 2; According to the calculation formula Obtain the operating status evaluation coefficient λ of each battery i , where Y′ represents the base value of voltage, σ i It is expressed as the correction factor corresponding to the voltage of each battery.

5. A method for predicting a battery open circuit voltage according to claim 4, characterized in that: The analysis results in a health status assessment coefficient for each battery. The specific analysis process is as follows: The rated capacity, rated internal resistance, rated number of cycles, current remaining available power, current internal resistance value, number of deep cycles and number of shallow cycles of each battery are recorded as ER i ,EN i 、EX i ,DR i DN i SX i and QX i ; According to the calculation formula Obtain the health status assessment coefficient β of each battery i , where ω1 and ω2 represent the weight factors corresponding to the number of deep cycles and the number of shallow cycles, respectively.

6. A method for predicting battery open circuit voltage according to claim 5, characterized in that: The specific process of judging the current operating status of each battery is as follows: The operating status evaluation coefficient of each battery is compared with the threshold of the operating status evaluation coefficient in the database. When the operating status evaluation coefficient of a battery is equal to the threshold of the operating status evaluation coefficient in the database, the operating status of the battery is judged to be stationary; when the operating status evaluation coefficient of a battery is less than the threshold of the operating status evaluation coefficient in the database, the operating status of the battery is judged to be charging; when the operating status evaluation coefficient of a battery is greater than the threshold of the operating status evaluation coefficient in the database, the operating status of the battery is judged to be discharging.

7. A method for predicting battery open circuit voltage according to claim 6, characterized in that: The open circuit voltage prediction method includes a basic prediction method, a data driven prediction method and a real-time prediction method.

8. A method for predicting battery open circuit voltage according to claim 7, characterized in that: The open circuit voltage prediction method of each battery is allocated, and the specific allocation process is as follows: When the operating state of a battery is static, the basic prediction method is used as the open circuit voltage prediction method of the battery; when the operating state of a battery is charging or discharging, the final voltage of the battery is compared with the predicted voltage threshold in the database. When the final voltage of the battery is less than or equal to the predicted voltage threshold in the database, the data-driven prediction method is used to analyze and obtain the battery voltage. When the final voltage of the battery is greater than the predicted voltage threshold in the database, the real-time prediction method is used to analyze and obtain the battery voltage.

9. A method for predicting battery open circuit voltage according to claim 8, characterized in that: According to the open circuit voltage prediction method of each battery, the open circuit voltage of each battery is predicted, and the specific prediction process is as follows: A1. Basic prediction method, according to the calculation formula Get the SOC of each battery i , where ρ i The correction factor of the health status assessment coefficient of each battery on the rated capacity of the battery is expressed as the correction factor, and according to the state of charge of each battery, in the functional relationship between the state of charge of each battery and the open circuit voltage, the open circuit voltage value corresponding to the state of charge of each battery is obtained, and the open circuit voltage value is used as the battery open circuit voltage of each battery; A2. Data-driven prediction method, firstly, the current internal resistance value and current temperature of each battery are transmitted to the relationship model of internal resistance, temperature and state of charge, and then the state of charge of each battery is output, so as to obtain the open circuit voltage value corresponding to the state of charge of each battery in the functional relationship between the state of charge of each battery and the open circuit voltage, and the open circuit voltage value is used as the battery open circuit voltage of each battery; A3. Real-time prediction method, according to the calculation formula: The battery state of charge is analyzed, where CSOC i Represents the initial state of charge of each battery, CSOC i Expressed as a percentage of the rated capacity of each battery, I i Expressed as the current value of each battery, t i It is expressed as the working time of each battery, so that according to the charge state of each battery, in the functional relationship between the charge state of each battery and the open circuit voltage, the open circuit voltage value corresponding to the charge state of each battery is obtained, and the open circuit voltage value is used as the battery open circuit voltage of each battery.

10. A method for predicting battery open circuit voltage according to claim 9, characterized in that: The relationship model between temperature and state of charge, the specific acquisition process is as follows: The temperature sensor and the internal resistance meter are used as data acquisition devices; the temperature and the state of charge are randomly selected as a set of experimental conditions in the temperature set and the state of charge set, thereby obtaining different sets of experimental conditions, and the battery is tested according to each set of experimental conditions, and then the internal resistance value of the battery under each set of experimental conditions is recorded during the experiment, and the temperature, state of charge and internal resistance value are used as a data set to obtain each data set; the data is checked, and the data sets containing abnormal values ​​or erroneous data are deleted, and then the neural network of the nonlinear regression model is used to establish the model; first, the data set is divided, and each data set is divided into a training set and a test set, and the training set data is used to train the selected model. During the training process, the parameters of the model are adjusted to optimize the performance, and the test set data is used to verify the performance of the model. The parameters of the model are optimized through the cross-validation method, thereby obtaining the relationship model between the internal resistance, temperature and state of charge.

Images