Battery life prediction method and device, nonvolatile storage medium and electronic equipment

Abstract

The invention discloses a battery life prediction method and device, a nonvolatile storage medium and electronic equipment. The method comprises the steps that cycle data corresponding to a reference battery in multiple different cycle stages are obtained, and the cycle data represent the corresponding relation between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stages; based on the cycle data corresponding to the plurality of cycle stages, determining a correlation coefficient representing an association relationship between the discharge capacity attenuation and the battery life; according to the characteristic parameters corresponding to the to-be-tested battery, a target fitting model is adopted, the service life prediction result of the to-be-tested battery is obtained, a battery system of the to-be-tested battery is matched with a battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient. The technical problem that the battery life prediction efficiency is not ideal in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of battery technology, and in particular to a battery life prediction method, device, non-volatile storage medium and electronic device. Background Art

[0002] Secondary batteries have been widely used in consumer electronics, electric vehicles, energy storage systems and other fields due to their high energy density, long cycle life and environmental friendliness. The life prediction of secondary batteries is a key technical problem. Accurately predicting the remaining life of secondary batteries can not only improve the utilization efficiency of batteries, but also avoid a series of safety issues caused by the degradation of battery performance.

[0003] The secondary battery life prediction methods in the related art are mainly based on empirical models and physical models. The empirical model usually relies on historical data and predicts the remaining life of the battery through statistical analysis methods. For example, by fitting the capacity decay curve, the remaining life of the battery under specific conditions can be inferred. However, this method has poor applicability in new battery systems and has high requirements for data volume and data quality. The physical model is based on the electrochemical reaction mechanism inside the battery, and predicts the performance changes of the battery through mathematical modeling and simulation. This method requires an in-depth understanding of the internal structure and reaction mechanism of the battery. The modeling process is complex and the computational cost is high. Therefore, there is a problem of unsatisfactory efficiency in battery life prediction in the related art.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present application provide a battery life prediction method, device, non-volatile storage medium and electronic device to at least solve the technical problem of unsatisfactory battery life prediction efficiency existing in the related art.

[0006] According to one aspect of an embodiment of the present application, a battery life prediction method is provided, including: obtaining cycle data corresponding to a reference battery at different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery at the corresponding cycle stage; determining the correlation coefficient representing the correlation between the discharge capacity decay and the battery life based on the cycle data corresponding to the multiple cycle stages; according to the characteristic parameters corresponding to the battery to be tested, using a target fitting model to obtain the life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient. Through the above processing, mathematical features are extracted from the discharge curves of different cycle stages and life prediction is performed, which avoids the need for full life cycle testing and significantly shortens the time required to determine the battery life. The use of data-driven thinking can help optimize the design and production process and reduce unnecessary testing costs and time.

[0007] Optionally, based on the cycle data corresponding to the multiple cycle stages, the correlation coefficient representing the correlation between the discharge capacity decay and the battery life is determined, including: determining the characteristic parameters representing the discharge capacity decay of the reference battery in the multiple cycle stages based on the cycle data corresponding to the multiple cycle stages; determining other batteries that match the battery system of the reference battery; determining the characteristic parameters of other batteries by determining the characteristic parameters of the reference battery; based on the characteristic parameters and battery life corresponding to the reference battery, and the characteristic parameters and battery life corresponding to the other batteries, determining the correlation coefficient representing the correlation between the discharge capacity decay and the battery life. Through the above processing, the characteristic parameters of multiple batteries with the same system at different cycle stages are calculated to ensure that the model can accurately capture the correlation between the battery life and the discharge curve characteristics.

[0008] Optionally, the method further includes: determining the initial total discharge capacity of the reference battery; determining the target discharge capacity based on the initial total discharge capacity and the attenuation threshold representing the battery life limit; and determining the number of cycles corresponding to the target discharge capacity based on the cycle data corresponding to the multiple cycle stages of the reference battery as the battery life of the reference battery. Through the above processing, the influence of the battery system on the life determination is reduced, the consistency of the evaluation process is ensured, and it is conducive to the comparison between different batteries and the training of the prediction model.

[0009] Optionally, based on the cycle data corresponding to the multiple cycle stages, characteristic parameters representing the discharge capacity decay of the reference battery in the multiple cycle stages are determined, including: based on the cycle data corresponding to the multiple cycle stages, according to the predetermined discharge voltage change interval, a predetermined number of data points are selected in the multiple cycle stages; the capacity difference corresponding to the predetermined number of data points at the same discharge voltage is determined; based on the capacity difference corresponding to the predetermined number of data points, a predetermined statistical method is used to process and obtain the characteristic parameters. Through the above processing, the predefined discharge voltage change interval and the number of data points are selected, and the data processing process is standardized, ensuring the data comparability and analysis consistency between different batteries or different cycle stages.

[0010] Optionally, the method further includes: determining the characteristics of the change in electrical performance of the reference battery; based on the characteristics of the change in electrical performance, dividing the total cycle data of the reference battery into stages to obtain cycle stage division results; based on the cycle stage division results, determining the first cycle stage in all cycle stages indicating that the rate of change in the electrical performance of the reference battery is less than a predetermined change threshold, and determining the second cycle stage indicating that the amplitude of the decrease in the electrical performance of the reference battery is less than a predetermined amplitude threshold, as multiple cycle stages. By analyzing and dividing the characteristics of the change in electrical performance of the reference battery into stages, the influence of the data of the severe degradation stage in the late life cycle on the accuracy of model prediction is avoided, the noise in the model training process can be reduced, and the stability and reliability of the prediction model can be improved.

[0011] Optionally, the method further includes: determining an initial fitting model from a plurality of candidate fitting models based on the correlation coefficient; and training the initial fitting model using historical discharge data matching the battery system of the reference battery to obtain a target fitting model. Through the above processing, the error caused by insufficient generalization of the model is reduced, and the performance change law within the battery system can be more accurately captured, thereby improving the accuracy of battery life prediction.

[0012] Optionally, the multiple candidate fitting models include a linear fitting model and a nonlinear fitting model, and based on the correlation coefficient, the initial fitting model is determined from the multiple candidate fitting models, including: determining the correlation coefficient ranges corresponding to the multiple candidate fitting models; when the correlation coefficient is in a first range, determining the initial fitting model to be a linear fitting model; when the correlation coefficient is in a second range, determining the initial fitting model to be a nonlinear fitting model, wherein the correlation represented by the first range is greater than the correlation represented by the second range. By analyzing the correlation coefficient, a choice can be made between the linear fitting model and the nonlinear fitting model, thereby improving the accuracy of the model fitting to the data.

[0013] According to another aspect of an embodiment of the present application, a battery life prediction device is provided, including: a cycle data acquisition module, used to obtain cycle data corresponding to a reference battery in different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage; a correlation determination module, used to determine a correlation coefficient representing the correlation between discharge capacity decay and battery life based on the cycle data corresponding to the multiple cycle stages; a life prediction module, used to obtain a life prediction result of the battery to be tested by using a target fitting model based on characteristic parameters corresponding to the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

[0014] According to another aspect of an embodiment of the present application, a battery life prediction system is provided, comprising: a battery cycling device, and a data processing device, wherein the battery cycling device is used to obtain cycle data corresponding to a reference battery in different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage; the data processing device is connected to the battery cycling device, and is used to determine a correlation coefficient representing the correlation between discharge capacity decay and battery life based on the cycle data corresponding to the multiple cycle stages; based on the characteristic parameters corresponding to the battery to be tested, a target fitting model is used to obtain a life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

[0015] According to another aspect of an embodiment of the present application, an electronic device is provided, comprising: one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by one or more processors, the one or more processors implement any one of the battery life prediction methods.

[0016] In the embodiment of the present application, by obtaining the cycle data corresponding to the reference battery in different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage; based on the cycle data corresponding to the multiple cycle stages, the correlation coefficient representing the correlation between the discharge capacity decay and the battery life is determined; according to the characteristic parameters corresponding to the battery to be tested, the target fitting model is adopted to obtain the life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient. The purpose of extracting discharge characteristics for battery life prediction is achieved, and the technical effect of improving the efficiency of battery life prediction is achieved, thereby solving the technical problem of unsatisfactory battery life prediction efficiency existing in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 is a flowchart of an optional battery life prediction method provided according to an embodiment of the present application;

[0019] Figure 2 is a structural block diagram of an optional battery life prediction system provided according to an embodiment of the present application;

[0020] Figure 3 is a schematic flow chart of an optional battery life prediction method provided according to an embodiment of the present application;

[0021] Figure 4 is a rendering of an optional battery life prediction method provided according to an embodiment of the present application;

[0022] Figure 5 It is a schematic diagram of an optional battery life prediction device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present application.

[0024] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0025] For the convenience of description, some nouns or terms involved in the embodiments of the present application are explained below:

[0026] Half-Cell, also known as half-cell, is a small battery used for preliminary testing of materials and formulations in the early stages of research and development. This type of battery usually contains only one active electrode (positive or negative electrode) and one counter electrode (such as metallic lithium as a reference electrode). The main purpose of half-cell design is to simplify the testing process and better study and analyze the performance of a single electrode, such as electrochemical stability, charge and discharge efficiency, cycle life, etc. Since only one active electrode in the half-cell is involved in the reaction, they can provide direct information about a specific material or formulation without being affected by the performance of the other electrode material.

[0027] Full-Cell refers to a complete battery system that includes positive and negative electrodes. Its design and working principle are closer to the final battery form. Full-cell testing can reflect the overall performance of the battery in actual use, including the matching between the positive and negative electrodes, the total energy density of the battery, cycle performance, and safety. Full-cell testing is a further performance evaluation based on half-cell testing to ensure the performance and stability of the battery in actual applications.

[0028] The life degradation of secondary batteries mainly includes two aspects: capacity decay and internal resistance increase. Capacity decay refers to the decrease in the number of active ions due to irreversible chemical reactions during the charge and discharge process, which gradually reduces the capacity of the battery; the increase in internal resistance is due to the aging of electrode materials, the degradation of electrolytes and the formation of interface films, which increases the internal resistance of the battery and affects the charge and discharge performance of the battery. Understanding these two degradation mechanisms is crucial for predicting the life of secondary batteries.

[0029] The empirical models in related technologies are highly dependent on a large amount of historical data and are mainly targeted at mature battery systems with good consistency. For new or early-stage solid-state battery systems, the accuracy and applicability of these models are limited due to the lack of sufficient data.

[0030] Although physical models based on electrochemical principles can deeply analyze the internal reaction mechanism of the battery, their construction and parameterization are very complex and require precise knowledge of the internal structure of the battery and high computational costs, which are often not feasible in practical applications. Due to factors such as the interface contact impedance between the solid electrolyte and the electrode material and temperature sensitivity, the process of capacity decay and internal resistance increase of solid-state batteries may show complex variation characteristics, making it difficult to accurately capture the complexity of battery life degradation.

[0031] In response to the above problems, an embodiment of the present application provides a method embodiment for battery life prediction. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0032] Figure 1 is a flowchart of a battery life prediction method according to an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:

[0033] Step S102, obtaining cycle data corresponding to a reference battery at different multiple cycle stages, wherein the cycle data represents a corresponding relationship between a discharge voltage and a discharge capacity of the reference battery at a corresponding cycle stage;

[0034] It can be understood that the collection of cycle data of reference batteries at different cycle stages is used to reflect the corresponding relationship between discharge voltage and discharge capacity, providing a basis for subsequent analysis. By extracting mathematical features from early (also called low-cycle) or mid-to-late (also called high-cycle) discharge curves and performing life prediction, the need for long-term cycle testing that relies on the entire life cycle of the battery is avoided, significantly shortening the prediction time while ensuring the accuracy of the prediction.

[0035] Optionally, the reference battery may be a solid-state battery, or a lithium-ion battery with a liquid electrolyte, etc.

[0036] In an optional embodiment, the method further includes: determining the electrical performance change characteristics of the reference battery; based on the electrical performance change characteristics, dividing the total cycle data of the reference battery into stages to obtain cycle stage division results; based on the cycle stage division results, determining, among all cycle stages, a first cycle stage indicating that the rate of change of the electrical performance of the reference battery is less than a predetermined change threshold, and determining a second cycle stage indicating that the amplitude of the electrical performance decrease of the reference battery is less than a predetermined amplitude threshold, as multiple cycle stages.

[0037] It can be understood that in order to select the cycle data of the reference battery for life prediction, the total cycle data of the reference battery is divided into stages by analyzing the characteristics of the change in the electrical performance of the battery. It is intended to identify the stable stage before the battery performance degradation, and the stage where the performance begins to decline significantly. First, the trend of the change in the electrical performance of the reference battery with the number of cycles is analyzed. Based on the above-mentioned characteristics of the change in electrical performance, the total cycle data of the reference battery can be divided into stages. Each stage represents the electrical performance state of the battery within a specific range of cycles. The purpose of stage division is to isolate the different stages in the battery performance degradation process so that representative data can be selected for subsequent analysis and model training. In the first cycle stage, the rate of change of electrical performance is less than the predetermined change threshold, which means that the performance of the battery is relatively stable and has not yet begun a significant degradation process. This is usually the early stage of the battery cycle. In the second cycle stage, the decline in electrical performance is less than the predetermined amplitude threshold, indicating that the battery performance begins to decline, but has not yet reached the stage of sharp degradation or the end of life, indicating the middle or middle and late stages of the battery cycle.

[0038] By analyzing and dividing the characteristics of the electrical performance changes of the reference battery into stages, the data for life prediction can be selected more finely. Only the data from the stage when the battery performance is stable and begins to slowly degrade is used to avoid the impact of the data from the drastic degradation stage in the later stage of the life cycle on the accuracy of model prediction. Selecting stage data with relatively stable electrical performance change rate and decline for modeling can reduce noise in the model training process and improve the stability and reliability of the prediction model.

[0039] Optionally, the stable working state of a battery refers to the state in which the rate of change of its electrochemical performance parameters (such as capacity, internal resistance, polarization, etc.) begins to slow down and enter a relatively stable state after a certain number of cycles after the battery is produced. The capacity of the battery will change after the initial few cycles due to factors such as activation and SEI film formation, and then the rate of change gradually decreases. A threshold value can be defined, such as a capacity change rate less than a certain percentage (such as within 1%), as the starting point of the stable state. Around the 50th cycle, the capacity change of many battery types (such as lithium-ion batteries) usually slows down significantly, indicating that the battery may have entered a stable state. The polarization of the battery, that is, the nonlinear relationship between the voltage and current of the battery during the charge and discharge process, will be more obvious in the initial cycle of the battery, and then gradually stabilize. The degree of polarization can also be quantified by its rate of change. When the rate of change of the degree of polarization decreases below a certain value, it can be considered that the battery has entered a stable state.

[0040] In practical applications, determining the starting point of the battery stable state needs to be adjusted according to factors such as battery type, material, manufacturing process, and test conditions. For example, for some solid-state batteries, the stable state may start after fewer cycles, while for other battery types, more cycles may be required to achieve stability.

[0041] The first cycle stage is preferably the 1st to 80th charge and discharge cycles, and the data from one of them (such as the 50th cycle data) can be selected for subsequent data processing.

[0042] Similarly, the degree of electrical performance degradation may include a decrease in capacity retention, an increase in internal resistance, and so on. The battery capacity gradually decreases with the number of cycles, and the capacity retention rate is the ratio of the current capacity of the battery to the initial capacity. The second cycle stage is preferably the 80th-150th charge and discharge cycle, and the data from one of them (such as the 150th cycle data) can be selected for subsequent data processing. The above-mentioned preferred methods for the first cycle stage and the second cycle stage are based on the results of experimental observations and data analysis of specific solid-state battery systems. This selection point can be adjusted and optimized with the development of battery technology and changes in specific application scenarios.

[0043] Step S104, determining a correlation coefficient representing a correlation relationship between discharge capacity decay and battery life based on cycle data corresponding to the plurality of cycle stages;

[0044] It can be understood that based on the collected cycle data, the statistical correlation between the attenuation of the discharge capacity and the battery life, i.e., the correlation coefficient, is calculated. In this embodiment, the Pearson correlation coefficient can be used. By quantifying the degree of dependence between the characteristics of the discharge curve and the battery life, guidance is provided for the selection of the fitting method. There is no need to perform long-term life tests on each battery system, which reduces the number of batteries and experimental cycles required for the test, thereby reducing R&D and production costs.

[0045] In an optional embodiment, a correlation coefficient representing the correlation between discharge capacity decay and battery life is determined based on cycle data corresponding to multiple cycle stages, including: determining characteristic parameters representing the discharge capacity decay of a reference battery in multiple cycle stages based on cycle data corresponding to multiple cycle stages; determining other batteries that match the battery system of the reference battery; determining characteristic parameters of other batteries by using the method of determining the characteristic parameters of the reference battery; determining the correlation coefficient representing the correlation between discharge capacity decay and battery life based on the characteristic parameters and battery life corresponding to the reference battery, and the characteristic parameters and battery life corresponding to other batteries.

[0046] It can be understood that based on the collected cycle data of multiple cycle stages, characteristic parameters that can describe the discharge capacity attenuation characteristics of the battery during these stages are determined. Other batteries with the same battery system as the reference battery are selected to ensure the comparability of the data and the applicability of the model. The characteristic parameters of these similar batteries are calculated using the same method as the reference battery to ensure that all battery data are processed in a consistent manner, allowing effective comparison and analysis. Based on the characteristic parameters of the reference battery and other similar batteries and the corresponding battery life, the correlation coefficient between the discharge capacity attenuation and the battery life is calculated. This correlation coefficient will be used to guide the training and selection of the fitting model to ensure that the model can accurately capture the correlation between the battery life and the discharge curve characteristics.

[0047] By calculating the characteristic parameters of multiple batteries with the same system at different cycle stages, the relationship between discharge capacity decay and battery life can be analyzed more comprehensively, thereby improving the accuracy and stability of the prediction model. Selecting multiple battery samples, not limited to reference batteries, but also including other batteries with the same system, ensures that the model can be trained on a wider range of battery data, improving the model's prediction ability for new battery samples.

[0048] In an optional embodiment, the method also includes: determining an initial total discharge capacity of a reference battery; determining a target discharge capacity based on the initial total discharge capacity and a decay threshold representing a battery life limit; and determining the number of cycles corresponding to the target discharge capacity as the battery life of the reference battery based on cycle data corresponding to multiple cycle stages of the reference battery.

[0049] It is understandable that the total discharge capacity of the reference battery in a new state needs to be measured and recorded as an important benchmark for battery life assessment, namely the initial total discharge capacity. Based on the initial total discharge capacity and a preset attenuation threshold, which is the battery performance attenuated to a certain percentage of the initial total discharge capacity, such as 80% or 85%, the target discharge capacity is determined. When the battery's discharge capacity drops to this target discharge capacity, it can be considered that the battery life has reached its limit.

[0050] By setting the initial total discharge capacity and attenuation threshold, the battery life can be defined more accurately, avoiding errors caused by subjective judgment, making the life prediction more objective. The consistency of the evaluation process is ensured. Regardless of the battery system or type, the same attenuation threshold is used as the standard for the life limit, which is conducive to comparison between different batteries and training of prediction models.

[0051] In an optional embodiment, based on the cycle data corresponding to the multiple cycle stages, characteristic parameters representing the discharge capacity attenuation of the reference battery in the multiple cycle stages are determined, including: based on the cycle data corresponding to the multiple cycle stages, according to a predetermined discharge voltage change interval, a predetermined number of data points are selected in the multiple cycle stages; at the same discharge voltage, the capacity difference corresponding to the predetermined number of data points is determined; based on the capacity difference corresponding to the predetermined number of data points, a predetermined statistical method is used to perform processing to obtain the characteristic parameters.

[0052] It can be understood that in each cycle stage, a fixed number of data points are selected according to the predefined discharge voltage change interval, and the selected data points are evenly distributed in the voltage range, such as selecting 1000 data points at intervals within the range of 2-3V (volts), which is convenient for subsequent statistical analysis and feature extraction. For each selected data point, the discharge capacity difference at different cycle stages under the same voltage is calculated, such as based on the discharge curve between two cycles (such as low-order and high-order cycles), which helps to quantify the discreteness of the discharge curve in the early cycle stage and reflects the stability of the internal structure and performance of the battery. The above-calculated capacity difference is processed using a predetermined statistical method (such as calculating the variance of the absolute value of the capacity difference) to generate characteristic parameters representing the discharge capacity attenuation characteristics, reflecting the performance status of the battery at a specific cycle stage, and providing key input for the battery life prediction model.

[0053] By pre-defining the discharge voltage change interval and the number of data points, the data processing process is standardized to ensure data comparability and analysis consistency between different batteries or different cycle stages. The characteristic parameters (such as variance values) used in this embodiment can effectively reflect the stability of the discharge capacity of the battery at a specific cycle stage, and are sensitive indicators of internal structural changes and performance degradation of the battery, which helps to improve the accuracy of the prediction model.

[0054] Step S106, according to the characteristic parameters corresponding to the battery to be tested, a target fitting model is used to obtain a life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

[0055] It can be understood that the target fitting model (such as a polynomial regression model) is selected and trained using the determined correlation coefficient to predict the remaining life of the battery to be tested. The selection of the target fitting model is based on the characteristic parameters and life data of the known reference battery. When predicting, the characteristic parameters of the battery to be tested of the same system are input, which can quickly and accurately evaluate the battery life and reduce the resource and time costs of traditional prediction technologies.

[0056] It should be noted that for problems such as the interface contact impedance and temperature sensitivity of solid-state batteries, life predictions can be provided in the early stages of battery development, which is of great significance for optimizing solid-state battery design and accelerating the implementation of battery products.

[0057] In an optional embodiment, the method further includes: determining an initial fitting model from a plurality of candidate fitting models based on a correlation coefficient; and training the initial fitting model using historical discharge data matching a battery system of a reference battery to obtain a target fitting model.

[0058] It can be understood that selecting the initial fitting model from multiple preset candidate fitting models based on the correlation coefficient between discharge capacity decay and battery life involves the decision of model type, such as linear regression, polynomial regression or other more complex regression models, to establish a mathematical relationship between capacity decay characteristics and battery life. The selected initial fitting model is trained using the historical discharge data of batteries of the same system as the reference battery to optimize the model parameters. After training, the initial fitting model is converted into a more accurate target fitting model. The target fitting model can predict the life of batteries of the same system based on the input characteristic parameters (such as variance values), thereby providing a basis for battery performance evaluation and optimization design.

[0059] By selecting the initial fitting model based on the correlation coefficient and training it with historical battery data matching the system, a more personalized and optimized prediction model can be obtained. This model has a stronger prediction ability for a specific battery system, reduces the error caused by insufficient model generalization, and can more accurately capture the performance change rules within the battery system, thereby improving the accuracy of battery life prediction.

[0060] In an optional embodiment, multiple candidate fitting models include a linear fitting model and a nonlinear fitting model, and based on the correlation coefficient, an initial fitting model is determined among the multiple candidate fitting models, including: determining the correlation coefficient ranges corresponding to the multiple candidate fitting models respectively; when the correlation coefficient is in a first range, determining that the initial fitting model is a linear fitting model; when the correlation coefficient is in a second range, determining that the initial fitting model is a nonlinear fitting model, wherein the correlation represented by the first range is greater than the correlation represented by the second range.

[0061] It can be understood that defining two correlation coefficient ranges, a first range and a second range, represents correlations of different strengths. The first range generally represents a stronger correlation, while the second range represents a weaker correlation. Based on the magnitude of the correlation coefficient, it is determined whether to use a linear fitting model or a nonlinear fitting model. When the correlation coefficient falls within the first range, the linear fitting model is selected, and when the correlation coefficient falls within the second range, the nonlinear fitting model is selected.

[0062] By analyzing the correlation coefficient, we can choose between a linear fitting model and a nonlinear fitting model, which improves the accuracy of the model's fitting of the data. Linear models perform well when the correlation is strong, while nonlinear models are more suitable when the correlation is weak or the relationship is complex. This method ensures that the model can achieve the best prediction effect under different correlation conditions.

[0063] Optionally, nonlinear fitting and linear fitting are used to find the most appropriate mathematical function relationship based on the data set. Linear fitting is suitable for situations where the data presents a linear relationship, while nonlinear fitting is more flexible and can handle more complex relationships in the data. Linear fitting means that the model function can be expressed as a linear combination of variables, that is, the relationship between the model function and the input variables is a straight line or a plane relationship.

[0064] There are many ways of linear fitting, for example, simple linear regression: when there is only one predictor variable, the model can be expressed as y = a + bx, where x is the independent variable, y is the dependent variable, a is the intercept, and b is the slope. Multiple linear regression: when there are multiple predictor variables, different variables can be distinguished by numbering. The model can be expressed as, for example, y = a + b1x1 +b2x2 + … + bnxn. Ridge Regression: on the basis of multiple linear regression, a regularization term is introduced to prevent the model from overfitting, which is suitable for situations with a large number of features. LASSO regression (Least Absolute Shrinkage and Selection Operator): regularization is also added to multiple linear regression, but LASSO regression can achieve feature selection through L1 regularization, that is, some regression coefficients can be precisely set to zero, thereby simplifying the model.

[0065] There are many ways to fit nonlinear data. Nonlinear fitting allows a nonlinear relationship between the model function and the input variables, which makes it possible to fit more complex data patterns. For example, polynomial regression allows the model function to contain quadratic, cubic or higher terms of the input variables, such as y = a + b1x + b2x 2 + b3x 3 , which can be used to fit the curve trend in the data. Exponential regression: The model function contains an exponential function, such as y = ae bx , used to describe growth or decay phenomena. Power law regression: The fitted model function is a power law relationship, such as y = ax b , which is suitable for describing the scale invariance during battery degradation.

[0066] Optionally, when the above correlation coefficient is in the third range, the correlation in the third range is less than the correlation in the second range, and the correlation is considered to be very small, and it is difficult to obtain a good processing effect by fitting. The correlation coefficient in the third range can be processed using an artificial intelligence algorithm. There can be many artificial intelligence algorithms. For example, a neural network algorithm can handle very complex nonlinear relationships and approximate arbitrarily complex functional relationships through multi-layer nonlinear transformations. Support Vector Regression (SVR), based on the principle of support vector machines, can handle linear and nonlinear regression problems and achieve nonlinear fitting through kernel techniques. The decision tree model can approximate nonlinear relationships in the data through a branching structure, and is particularly suitable for situations where there are piecewise linear or piecewise constant relationships in the data. Gaussian Process Regression (GPR), a model based on probability theory, is used to fit nonlinear trends in data and can also provide uncertainty estimates for predictions.

[0067] In the field of battery life prediction, nonlinear fitting methods can capture the complex and nonlinear changes of electrochemical performance over time. Polynomial regression can be used to fit the nonlinear attenuation trend of discharge capacity with the number of cycles, while neural networks can learn the complex electrochemical reaction mechanism inside the battery to achieve more accurate life prediction.

[0068] Through the above step S102, the cycle data corresponding to the reference battery in different multiple cycle stages are obtained, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage; step S104, based on the cycle data corresponding to the multiple cycle stages, the correlation coefficient representing the correlation between the discharge capacity decay and the battery life is determined; step S106, according to the characteristic parameters corresponding to the battery to be tested, the target fitting model is used to obtain the life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient. The purpose of extracting discharge characteristics for battery life prediction is achieved, and the technical effect of improving the efficiency of battery life prediction is achieved, thereby solving the technical problem of unsatisfactory battery life prediction efficiency existing in the related art.

[0069] A battery life prediction system is also provided in an embodiment of the present application. The battery life prediction system provided in an embodiment of the present application is introduced below.

[0070] Figure 2 is a structural block diagram of a battery life prediction system provided according to an embodiment of the present application, such as Figure 2 As shown, the system includes: a battery circulation device 202 and a data processing device 204, and the description will be made based on this system.

[0071] A battery cycling device 202, used to obtain cycle data corresponding to a reference battery at different multiple cycle stages, wherein the cycle data represents a corresponding relationship between a discharge voltage and a discharge capacity of the reference battery at a corresponding cycle stage;

[0072] The data processing device 204 is connected to the battery cycling device 202, and is used to determine the correlation coefficient representing the correlation between the discharge capacity decay and the battery life based on the cycle data corresponding to the multiple cycle stages; according to the characteristic parameters corresponding to the battery to be tested, a target fitting model is used to obtain the life prediction result of the battery to be tested, wherein the battery system of the battery to be tested is matched with the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

[0073] In a battery life prediction system provided by an embodiment of the present application, a battery cycle device 202 is used to obtain cycle data corresponding to a reference battery in different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage; a data processing device 204 is connected to the battery cycle device 202, and is used to determine the correlation coefficient representing the correlation between the discharge capacity decay and the battery life based on the cycle data corresponding to the multiple cycle stages; based on the characteristic parameters corresponding to the battery to be tested, a target fitting model is used to obtain the life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient. The purpose of extracting discharge characteristics for battery life prediction is achieved, and the technical effect of improving the efficiency of battery life prediction is achieved, thereby solving the technical problem of unsatisfactory battery life prediction efficiency existing in the related art.

[0074] Based on the above embodiments and optional embodiments, the present application proposes an optional implementation mode: Figure 3 is a schematic flow chart of an optional battery life prediction method provided according to an embodiment of the present application, such as Figure 3 As shown, it is applied to the life prediction of solid-state batteries, as explained below.

[0075] Step S1, collects two discharge voltages (50th and 150th times) and discharge capacity data in the solid-state battery cycle, as well as the battery life. The two discharge curves in step S1 are for batteries of the same system and correspond to the same number of cycles.

[0076] The discharge voltage and discharge capacity of the new solid-state battery in two different charge and discharge cycles and the current battery life are collected to obtain the discharge curve corresponding to the two charge and discharge cycles; the two different charge and discharge cycles are specifically: the low number is one of the 1-80 charge and discharge cycles, and the high number is one of the 80-150 charge and discharge cycles. In specific implementation, the low number is the 50th charge and discharge cycle, and the high number is the 150th charge and discharge cycle. The life of the solid-state battery is obtained by experiment. The lithium battery is continuously charged and discharged. When the actual total capacity of the solid-state battery decays to 85% of the initial total capacity of the lithium battery, the current number of charge and discharge cycles is the life of the solid-state battery.

[0077] Step S2, under the same voltage range, batch process to obtain the capacity difference curves of all known solid-state batteries.

[0078] To normalize the voltage-capacity data, all discharge curves were fitted as spline functions and linearly interpolated. The capacity was fitted as a function of voltage and was fitted at 1000 linearly spaced voltage points within the same voltage range. These uniformly sized vectors support direct data operations such as subtraction, which can obtain capacity difference data at different voltage points and then calculate the variance.

[0079] The above processing method can optionally use Python algorithm for data processing. The capacity and voltage values ​​of the two discharge curves are input into the system, and the discharge voltage is used as the independent variable and the discharge capacity is used as the dependent variable. Interpolation fitting is performed in the range of 2-3V (volts), with an interval of 1000 points, to obtain 1000 capacity differences.

[0080] Step S3, calculate the variance of the absolute value of each battery capacity difference and count the battery life. Take the absolute value of the obtained capacity difference and calculate the variance. This represents the degree of dispersion of the discharge curve in two charge and discharge cycles.

[0081] Step S4, calculating the correlation coefficient between the variance value and the lifespan. In step S4, steps S1-S3 are repeated to perform the same processing on solid-state batteries of the same system to obtain the lifespan of each solid-state battery and the mathematical feature corresponding to each solid-state battery, namely the variance value.

[0082] For the variance value and life of solid-state batteries of the same system, one part is mold batteries and the other part is full batteries. Both groups of batteries use positive electrodes with different formulas, but there is still correlation.

[0083] The correlation coefficient is the Pearson correlation coefficient, which is calculated using the following formula:

[0084]

[0085] in, represents the correlation coefficient between the capacity variance corresponding to the same discharge voltage combination of all solid-state batteries and the life of the corresponding solid-state battery, X represents the variance set, Y represents the set of the life of all solid-state batteries, and E( ) represents the expected operation.

[0086] Step S5, training the solid-state battery life prediction regression model (i.e., the initial fitting model) based on the training set to obtain a trained lithium battery life prediction regression model (i.e., the target fitting model). The life prediction regression model selects a linear regression model or a nonlinear regression model based on the distribution relationship between the variance value and the life of the lithium battery.

[0087] Step S6: Input the variance value of the battery to be predicted under the corresponding voltage range into the corresponding regression model for prediction according to different systems, and output the life of the solid-state battery to be predicted. The correlation result of the mold battery pack is -0.8~-0.9, and the correlation of the full battery pack is about -0.7~0.8, so nonlinear regression fitting can be used, specifically polynomial regression.

[0088] Figure 4 is a rendering of an optional battery life prediction method provided in an embodiment of the present application, such as Figure 4 As shown, the fitting curves of mold batteries with different negative electrode materials (graphite, silicon carbon) and full batteries are combined respectively. The independent variable is the variance value (ie, the characteristic parameter) and the dependent variable is the lifespan.

[0089] Calculate the capacity difference corresponding to two different charge and discharge cycles of the lithium battery to be predicted, input the variance value to be predicted into the trained lithium battery life prediction regression model for prediction, and output the life of the current lithium battery to be predicted. According to specific actual tests, for solid-state batteries predicted by the above model, for mold batteries, full batteries, and using different positive and negative electrode materials, even if the formula is fine-tuned, the prediction accuracy remains within an acceptable range (error ≤ 15%), which shows that the model has good generalization ability.

[0090] The optional implementations described above achieve at least the following effects: by extracting mathematical features from early and mid-late discharge curves and performing life prediction, the need for full life cycle testing is avoided, significantly shortening the time required to determine battery life. The use of data-driven thinking can help optimize the design and production process and reduce unnecessary testing costs and time.

[0091] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0092] In this embodiment, a battery life prediction device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the terms "module" and "device" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0093] According to an embodiment of the present application, a device embodiment for implementing a battery life prediction method is also provided. Figure 5 is a schematic diagram of a battery life prediction device according to an embodiment of the present application, such as Figure 5 As shown, the battery life prediction device includes: a cycle data acquisition module 502, a correlation determination module 504, and a life prediction module 506. The device is described below.

[0094] The cycle data acquisition module 502 is used to acquire cycle data corresponding to the reference battery in different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage;

[0095] A correlation determination module 504, connected to the cycle data acquisition module 502, for determining a correlation coefficient representing a correlation relationship between discharge capacity decay and battery life based on cycle data corresponding to a plurality of cycle stages;

[0096] The life prediction module 506 is connected to the correlation determination module 504, and is used to obtain the life prediction result of the battery to be tested by using the target fitting model according to the characteristic parameters corresponding to the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

[0097] In a battery life prediction device provided by an embodiment of the present application, a cycle data acquisition module 502 is set to obtain the cycle data corresponding to the reference battery in different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage; a correlation determination module 504 is connected to the cycle data acquisition module 502, and is used to determine the correlation coefficient representing the correlation relationship between the discharge capacity decay and the battery life based on the cycle data corresponding to the multiple cycle stages; a life prediction module 506 is connected to the correlation determination module 504, and is used to obtain the life prediction result of the battery to be tested by using the target fitting model according to the characteristic parameters corresponding to the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient. The purpose of extracting discharge characteristics for battery life prediction is achieved, and the technical effect of improving the efficiency of battery life prediction is achieved, thereby solving the technical problem of unsatisfactory battery life prediction efficiency existing in the related art.

[0098] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0099] It should be noted that the above-mentioned cycle data acquisition module 502, correlation determination module 504, and life prediction module 506 correspond to steps S102 to S106 in the embodiment, and the examples and application scenarios implemented by the above-mentioned modules and corresponding steps are the same, but are not limited to the contents disclosed in the above-mentioned embodiment. It should be noted that the above-mentioned modules as part of the device can be run in a computer terminal.

[0100] It should be noted that the optional or preferred implementation of this embodiment can refer to the relevant description in the embodiment, which will not be repeated here.

[0101] The above-mentioned battery life prediction device may also include a processor and a memory. The cycle data acquisition module 502, the correlation determination module 504, the life prediction module 506, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions.

[0102] The processor includes a kernel, which retrieves the corresponding program unit from the memory. There can be one or more kernels. The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0103] The embodiment of the present application provides an electronic device, which includes a processor, a memory, and a program stored in the memory and can be run on the processor. When the processor executes the program, the following steps are implemented: obtaining cycle data corresponding to a reference battery in different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage; determining the correlation coefficient representing the correlation between the discharge capacity decay and the battery life based on the cycle data corresponding to the multiple cycle stages; according to the characteristic parameters corresponding to the battery to be tested, using the target fitting model, obtaining the life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient. The device in this article can be a server, a PC, etc.

[0104] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that initializes the following method steps: obtaining cycle data corresponding to a reference battery in multiple different cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery in the corresponding cycle stage; determining a correlation coefficient representing the correlation between discharge capacity attenuation and battery life based on the cycle data corresponding to the multiple cycle stages; and obtaining a life prediction result of the battery to be tested by using a target fitting model based on characteristic parameters corresponding to the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

[0105] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0106] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0107] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0109] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0110] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0111] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0112] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0113] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0114] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A battery life prediction method, characterized in that: include: Acquire cycle data corresponding to a reference battery at different multiple cycle stages, wherein the cycle data represents a corresponding relationship between a discharge voltage and a discharge capacity of the reference battery at a corresponding cycle stage; Determine a correlation coefficient representing a correlation between discharge capacity decay and battery life based on cycle data corresponding to each of the plurality of cycle stages; According to the characteristic parameters corresponding to the battery to be tested, a target fitting model is adopted to obtain a life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

2. The method according to claim 1, characterized in that The determining of the correlation coefficient indicating the correlation between the discharge capacity decay and the battery life based on the cycle data corresponding to the plurality of cycle stages respectively includes: Determining characteristic parameters representing discharge capacity decay of the reference battery in the multiple cycle stages based on the cycle data corresponding to the multiple cycle stages respectively; Determining other batteries that match the battery system of the reference battery; Determine the characteristic parameters of the other batteries by using the method of determining the characteristic parameters of the reference battery; Based on the characteristic parameters and battery life corresponding to the reference battery and the characteristic parameters and battery life corresponding to the other batteries, a correlation coefficient representing a correlation relationship between discharge capacity decay and battery life is determined.

3. The method according to claim 2, characterized in that The method further comprises: determining an initial total discharge capacity of the reference battery; Determining a target discharge capacity based on the initial total discharge capacity and a decay threshold representing a battery life limit; Based on the cycle data respectively corresponding to the multiple cycle stages of the reference battery, the number of cycles corresponding to the target discharge capacity is determined as the battery life of the reference battery.

4. The method according to claim 2, characterized in that: The determining, based on the cycle data corresponding to the multiple cycle stages respectively, characteristic parameters representing the discharge capacity decay of the reference battery in the multiple cycle stages comprises: Based on the cycle data corresponding to the multiple cycle stages respectively, selecting a predetermined number of data points in the multiple cycle stages respectively according to a predetermined discharge voltage change interval; Determine the capacity differences corresponding to the predetermined number of data points at the same discharge voltage; Based on the capacity difference values ​​respectively corresponding to the predetermined number of data points, a predetermined statistical method is adopted to perform processing to obtain the characteristic parameter.

5. The method according to claim 1, characterized in that The method further comprises: Determining the electrical performance variation characteristics of the reference battery; Based on the electrical performance change characteristics, the total cycle data of the reference battery is divided into stages to obtain a cycle stage division result; Based on the cycle stage division result, a first cycle stage representing that the rate of change of the electrical performance of the reference battery is less than a predetermined change threshold value is determined in all cycle stages, and a second cycle stage representing that the amplitude of the decrease in the electrical performance of the reference battery is less than a predetermined amplitude threshold value is determined as the multiple cycle stages.

6. The method according to any one of claims 1 to 5, characterized in that The method further comprises: Based on the correlation coefficient, determining an initial fitting model from a plurality of candidate fitting models; The initial fitting model is trained using historical discharge data that matches the battery system of the reference battery to obtain the target fitting model.

7. The method according to claim 6, characterized in that The multiple candidate fitting models include a linear fitting model and a nonlinear fitting model, and determining an initial fitting model from the multiple candidate fitting models based on the correlation coefficient includes: Determine the correlation coefficient ranges corresponding to the multiple candidate fitting models respectively; When the correlation coefficient is within the first range, determining that the initial fitting model is a linear fitting model; When the correlation coefficient is in the second range, the initial fitting model is determined to be the nonlinear fitting model, wherein the correlation represented by the first range is greater than the correlation represented by the second range.

8. A battery life prediction device, characterized in that: include: A cycle data acquisition module, used to acquire cycle data corresponding to a reference battery at different multiple cycle stages, wherein the cycle data represents a corresponding relationship between a discharge voltage and a discharge capacity of the reference battery at a corresponding cycle stage; A correlation determination module, for determining a correlation coefficient representing a correlation relationship between discharge capacity decay and battery life based on cycle data corresponding to a plurality of cycle stages; The life prediction module is used to obtain the life prediction result of the battery to be tested by adopting a target fitting model according to the characteristic parameters corresponding to the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

9. A battery life prediction system, characterized in that: include: Battery cycling equipment, and data processing equipment, wherein The battery cycling device is used to obtain cycle data corresponding to a reference battery at different multiple cycle stages, wherein the cycle data represents the corresponding relationship between the discharge voltage and the discharge capacity of the reference battery at the corresponding cycle stage; The data processing device is connected to the battery cycling device, and is used to determine the correlation coefficient representing the correlation between the discharge capacity decay and the battery life based on the cycle data corresponding to the multiple cycle stages; according to the characteristic parameters corresponding to the battery to be tested, a target fitting model is used to obtain the life prediction result of the battery to be tested, wherein the battery system of the battery to be tested matches the battery system of the reference battery, and the target fitting model is determined based on the correlation coefficient.

10. An electronic device, characterized in that: include: One or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the battery life prediction method described in any one of claims 1 to 7.

Images