Energy storage device life prediction method and system, electronic device, and storage medium

Abstract

The present application provides a method and system for predicting the life of an energy storage device, an electronic device, and a storage medium, and belongs to the technical field of equipment health prediction. The method includes: obtaining operating data corresponding to a target energy storage device; obtaining a target correlation index, which is a correlation index between multiple factors and the life of the energy storage device, and the target correlation index is calculated based on the historical operating data of multiple energy storage devices; obtaining a target nonlinear change characteristic, which is a nonlinear change characteristic between a multi-factor correlation vector extracted based on a causal index and the life of the energy storage device, and the multi-factor correlation vector is constructed based on data corresponding to multiple factors; inputting the target correlation index, the target nonlinear change characteristic, and the operating data corresponding to the target energy storage device in the target time period into an energy storage device life prediction model to perform equipment life prediction and obtain a life prediction result. The present application can improve the accuracy of energy storage device life prediction.

Description

Technical Field

[0001] The present application belongs to the technical field of equipment health prediction, and more specifically, relates to a method and system for predicting the life of energy storage equipment, electronic equipment, and storage media. Background Art

[0002] In the energy storage sector, accurately predicting the lifespan of energy storage equipment is crucial for optimizing system design, ensuring power supply reliability, and reducing operation and maintenance costs. In reality, the aging and lifespan of energy storage equipment are not only related to the duration or frequency of use, but are also affected by a variety of complex factors.

[0003] Currently, traditional energy storage device lifespan prediction methods have shortcomings. For one thing, some methods based on empirical formulas are overly rigid, resulting in poor adaptability and accuracy. Furthermore, some existing technologies rely on traditional machine learning algorithms such as linear regression or support vector machines to estimate device lifespan. However, these methods struggle to fully capture the complex coupling effects of the multiple factors that influence device lifespan, resulting in low prediction accuracy. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for predicting the life of an energy storage device, an electronic device, and a storage medium to improve the accuracy of energy storage device life prediction.

[0005] A first aspect of an embodiment of the present application provides a method for predicting the life of an energy storage device, comprising:

[0006] Obtaining operating data corresponding to a target energy storage device within a target time period, wherein the target energy storage device is the energy storage device to be predicted;

[0007] Obtaining a target correlation index, where the target correlation index is a correlation index between multiple factors and the life of the energy storage device, the multiple factors including ambient temperature, charge and discharge rate, charge and discharge depth, and usage time, and the target correlation index is calculated based on historical operating data of the multiple energy storage devices;

[0008] Obtaining a target nonlinear change characteristic, wherein the target nonlinear change characteristic is a nonlinear change characteristic between a multi-factor correlation vector extracted based on a causal index and the life of the energy storage device, wherein the causal index is a causal index between every two factors among the multiple factors, and the multi-factor correlation vector is constructed based on data corresponding to the multiple factors; the data corresponding to the multiple factors include ambient temperature data, charge and discharge rate data, charge and discharge depth data, and usage time data;

[0009] The target correlation index, the target nonlinear change characteristics and the operating data corresponding to the target energy storage device in the target time period are input into the energy storage device life prediction model to perform device life prediction and obtain a life prediction result.

[0010] A second aspect of an embodiment of the present application provides a system for predicting the life of an energy storage device, comprising:

[0011] An operation data acquisition module is used to acquire operation data corresponding to a target energy storage device within a target time period, wherein the target energy storage device is the energy storage device to be predicted;

[0012] a correlation index acquisition module, configured to acquire a target correlation index, wherein the target correlation index is a correlation index between multiple factors and the life of the energy storage device, the multiple factors including ambient temperature, charge and discharge rate, charge and discharge depth, and usage time, and the target correlation index is calculated based on historical operating data of multiple energy storage devices;

[0013] a nonlinear feature acquisition module, configured to acquire a target nonlinear change feature, wherein the target nonlinear change feature is a nonlinear change feature between a multi-factor association vector extracted based on a causal index and the life of the energy storage device, wherein the causal index is a causal index between every two factors among the multiple factors, and the multi-factor association vector is constructed based on data corresponding to the multiple factors; the data corresponding to the multiple factors includes ambient temperature data, charge and discharge rate data, charge and discharge depth data, and usage time data;

[0014] The life prediction module is used to input the target correlation index, the target nonlinear change characteristics and the operating data corresponding to the target energy storage device in the target time period into the energy storage device life prediction model to perform equipment life prediction and obtain a life prediction result.

[0015] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the above-mentioned energy storage device life prediction method when executing the computer program.

[0016] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned energy storage device life prediction method are implemented.

[0017] The beneficial effects of the energy storage device life prediction method and system, electronic device, and storage medium provided by the embodiments of the present application are: by calculating the correlation index between multiple factors and the life of the energy storage device, the embodiments of the present application can accurately screen out the key factors that have a significant impact on the life, avoid redundant information interference, and lay a reliable foundation for prediction. The embodiments of the present application extract nonlinear change characteristics based on the causal index, which can deeply capture the complex interactions and nonlinear relationships between multiple factors, and effectively make up for the shortcomings of traditional methods in nonlinear modeling. The embodiments of the present application input the correlation index, nonlinear change characteristics, and operating data into the prediction model to achieve a more comprehensive and accurate characterization of the life of the energy storage device. Compared with traditional methods, it significantly improves the prediction accuracy, helps to optimize the operation and maintenance strategy of the energy storage equipment, and ensures the stable and efficient operation of the energy storage system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. 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 any creative work.

[0019] Figure 1 A flow chart of a method for predicting the life of an energy storage device provided in one embodiment of the present application;

[0020] Figure 2 A structural block diagram of an energy storage device life prediction system provided in one embodiment of the present application;

[0021] Figure 3 A schematic block diagram of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0022] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0023] In order to make the purpose, technical solutions and advantages of this application clearer, specific embodiments will be described below with reference to the accompanying drawings.

[0024] Please refer to Figure 1 , Figure 1This is a flow chart of a method for predicting the life of an energy storage device provided in one embodiment of the present application. The method can be executed by an electronic device. Specifically, the method can include S101 to S104.

[0025] S101: Acquire operating data corresponding to a target energy storage device within a target time period, where the target energy storage device is the energy storage device to be predicted.

[0026] In this embodiment, the target time period refers to the pre-set time interval for the data generated by the operating data to be collected. The target time period can be flexibly set according to the forecast requirements. For example, short-term forecasts can use data from the past month, and long-term forecasts can use data from the past year. The target energy storage device is the specific energy storage device for which lifespan forecasting is required. The unique identifier of the device can be determined by parameters such as the device number, model, and installation location. The operating data can include ambient temperature data, charge and discharge rate data, charge and discharge depth data, usage time data, and basic electrical data such as battery voltage, current, and internal impedance.

[0027] In this embodiment, the operating data also includes equipment operating status information, which is closely related to the equipment life. Obtaining this data is the basis for life prediction. Only by accurately grasping the operating status of the equipment in a specific time period can we combine the correlation index and nonlinear change characteristics to obtain accurate life prediction results through the prediction model.

[0028] For example, this embodiment can uniquely identify the target energy storage device by device ID, physical address, or asset code. This embodiment can utilize a pre-deployed sensor network, such as temperature sensors, current transformers, and voltage transmitters, to collect real-time ambient temperature and charge / discharge current / voltage data. This embodiment can interface with a battery management system to obtain internal status data such as charge / discharge depth, state of charge (the percentage of the battery's remaining charge relative to its total capacity), and cycle count, in order to collect operational data corresponding to the target energy storage device.

[0029] This embodiment can set a target time period based on predicted demand, perform data cleaning on the operating data within the target time period, detect and replace abnormal values, and obtain accurate and complete operating data.

[0030] S102: Obtain a target correlation index. The target correlation index is a correlation index between multiple factors and the life of the energy storage device. The multiple factors include ambient temperature, charge and discharge rate, charge and discharge depth, and usage time. The target correlation index is calculated based on historical operating data of multiple energy storage devices.

[0031] In this embodiment, the target correlation index is a quantitative indicator that measures the degree of correlation between factors such as ambient temperature and charge / discharge rate and the lifespan of the energy storage device. The target correlation index may include the Pearson correlation coefficient between each factor and the lifespan. Historical operating data may include the operating parameters of multiple energy storage devices over a period of time and their corresponding lifespan indicators.

[0032] This example uses a correlation index to quantify the linear relationship between factors and lifespan, screening factors that significantly impact lifespan and assigning weights to them, providing key features for subsequent lifespan prediction. Highly correlated factors are important drivers of lifespan variations. For example, a strong negative correlation between charge and discharge rate and lifespan indicates that high rates accelerate device aging. Incorporating these factors into the lifespan prediction model can improve prediction accuracy.

[0033] For example, this embodiment can obtain historical operating data such as ambient temperature, charge and discharge rate, charge and discharge depth, and usage time from data sources such as battery management systems and sensors of multiple energy storage devices, and simultaneously obtain life-related data such as the corresponding health status and remaining number of cycles of the devices.

[0034] This embodiment cleans the acquired data, checks data integrity, deletes records with excessive missing values, fills in small amounts of missing data with adjacent values, and identifies and removes abnormal data, such as temperature values ​​or multiplier values ​​outside the normal range. This embodiment organizes the collected data by device and time sequence to ensure that the operating data and lifespan data of the same device are consistent over time.

[0035] This embodiment uses the Pearson correlation coefficient analysis method to calculate the target correlation index between each factor and the equipment life, providing a calculation basis for the life prediction of the energy storage equipment life prediction model.

[0036] S103: Obtaining a target nonlinear change characteristic, where the target nonlinear change characteristic is a nonlinear change characteristic between a multi-factor correlation vector extracted based on a causal index and the life of the energy storage device. The causal index is a causal index between every two factors among the multiple factors, and the multi-factor correlation vector is constructed based on data corresponding to the multiple factors; the data corresponding to the multiple factors include ambient temperature data, charge and discharge rate data, charge and discharge depth data, and usage time data.

[0037] In this embodiment, the target nonlinear change characteristic refers to the nonlinear impact pattern of the complex interactions between multiple factors on the lifespan of the energy storage device, as distinguished from the linear correlation of a single factor. For example, target nonlinear change characteristics may include: the combined effect of multiple factors on lifespan is greater or less than the sum of their individual effects; a factor has a minimal effect on lifespan within a specific threshold, but after exceeding the threshold, it interacts with other factors to cause a sudden decrease in lifespan; the effects of factors produce cumulative nonlinear effects over time; the presence of a factor weakens the effects of other factors on lifespan; and the impact of a factor on lifespan only becomes apparent after interacting with other factors with a time delay.

[0038] For example, when a high temperature environment is superimposed with a high charge and discharge rate, the battery life decay rate significantly exceeds the linear superposition of the two independent effects, presenting a nonlinear enhancement effect of "1+1>2". When the depth of charge and discharge is less than 80%, the ambient temperature has little effect on the life, but after exceeding 80%, the temperature increase will trigger a nonlinear inflection point where the life accelerates. Long-term low charge and discharge rate but high frequency cycle conditions cause progressive damage to the battery through time accumulation, and its life decay curve shows a nonlinear characteristic that accelerates with prolonged use. Good temperature control measures can suppress the negative impact of high charge and discharge depth on battery life, so that the decay rate when the two work together is lower than that of a single high depth condition.

[0039] A causal index is a quantitative measure of the strength and direction of causal relationships between multiple factors. This index can include the F statistic from a Granger causality test, the probability of a causal relationship (p-value), and causal effect values ​​in structural causal models, such as the average treatment effect (ATE). In this example, the causal index refers to the causal index between each two factors.

[0040] A multi-factor correlation vector is a multidimensional feature vector formed by aggregating the raw data and derived features of multiple factors over a time window. It is used to characterize the combined effects of multiple factors. Parameters of the multi-factor correlation vector can include ambient temperature (mean / maximum), charge / discharge rate (percentage of high rates), charge / discharge depth (mean / number of times exceeding 80%), usage time, and causal interaction terms such as temperature × rate and rate × depth.

[0041] In this embodiment, the lifespan of energy storage devices is affected by the nonlinear interaction of multiple factors, rather than the independent effects of any single factor. For example, the synergistic effect of high temperature and high rate significantly accelerates capacity decay compared to either alone. The timing dependence of charge and discharge depth on temperature, such as long-term deep discharge at high temperatures, can cause irreversible aging.

[0042] The causal index in this embodiment is used to identify key causal chains and guide feature construction. For example, the causal index can be used to screen for factor pairs with strong causal relationships, such as temperature → multiplication factor, with p < 0.05, to avoid incorporating spurious correlations into the prediction model. Furthermore, cross-terms are constructed only for factor pairs with causal relationships, ensuring that nonlinear features have physical meaning rather than blind combinations.

[0043] The multi-factor correlation vector of this embodiment can aggregate high-frequency operating data according to time windows and convert them into low-frequency feature vectors, which are aligned with the life indicator in the time dimension, making it easier to capture long-term accumulated nonlinear effects.

[0044] For example, this embodiment can perform a Granger causality test on each pair of factors, calculate the F statistic and p-value, and then determine whether there is a significant causal relationship by comparing the thresholds. For factor pairs with a significant causal relationship, the causal direction and causal index value are recorded, such as the size of the F statistic, which reflects the strength of the effect.

[0045] This embodiment can segment historical data by sliding into fixed time windows, generating a feature vector for each window. This embodiment calculates feature data within the window, such as the mean temperature, maximum magnification, mean depth, and total usage time. This embodiment can also calculate features such as the standard deviation of temperature within the window and the proportion of time with high magnification to capture temporal variation patterns. This embodiment can also construct product terms for factor pairs that are significant in causal tests to avoid meaningless feature combinations. This embodiment can construct a multi-factor association vector based on the extracted feature data and product terms.

[0046] In this embodiment, the execution order of the combination of S101 and S102 and S103 can be swapped, for example, executing S102, S103, S101, etc. Figure 1 This is just an example execution order.

[0047] This embodiment can use models such as long short-term memory networks or gradient boosting trees that are good at processing nonlinear relationships and time series data to implement the above-mentioned specific steps of "extracting nonlinear change characteristics between multi-factor correlation vectors and energy storage device life based on causal index".

[0048] S104: Inputting the target correlation index, the target nonlinear change characteristics, and the operating data corresponding to the target energy storage device in the target time period into the energy storage device life prediction model to perform device life prediction and obtain a life prediction result.

[0049] Considering that a linear model with a single factor cannot capture the complexity of energy storage device aging, a purely nonlinear model will ignore physical priors. Fusion of the two types of features can balance the accuracy and interpretability of the model. Specifically, the target correlation index of this embodiment can provide a direct impact path of a single factor, such as high rate directly leading to life attenuation. The target nonlinear change characteristics of this embodiment can be used to supplement the indirect effects of multi-factor interactions, such as the synergistic acceleration of high temperature and high rate. The combination of the two can comprehensively characterize the characteristics of life variation.

[0050] This embodiment uses historical data training models to learn the mapping relationship from correlation index, nonlinear characteristics, real-time operation data to equipment life, thereby achieving accurate assessment of the current status of the equipment.

[0051] Exemplarily, the energy storage equipment life prediction model can adopt a fully connected neural network model, and the model can be trained through historical data. Specifically, it can include: training the initial model through multiple training samples to obtain the energy storage equipment life prediction model, wherein a training sample includes historical data and the life label corresponding to the historical data (such as health status level, remaining life duration, etc.), and the historical data can include historical correlation index, historical nonlinear change characteristics, and historical operation data. Furthermore, the historical data is used to predict the energy storage equipment life through the initial model to obtain a life prediction result. Based on the difference between the life prediction result and the corresponding life label, the initial model is trained to obtain the energy storage equipment life prediction model.

[0052] For example, the energy storage device life prediction model of this embodiment can also employ a multivariate linear regression model, using the historical correlation index, historical nonlinear feature vectors, and historical operating data statistics as the input feature matrix X, with dimensions [n × m], where n is the number of samples and m is the number of features. The weight vector W is solved using the least squares method to minimize the mean squared error between the predicted life value Y' = X × W and the actual life label Y, thereby further developing the energy storage device life prediction model.

[0053] For example, this embodiment uses statistical methods to calculate the correlation index between each factor and equipment lifespan, obtaining a quantitative value for the degree of association between each factor and lifespan. Factors are then weighted based on the absolute value of the correlation index. The larger the absolute value of the correlation index, the higher the weight of the corresponding factor. The raw data for each factor is multiplied by the corresponding weight and then added together to generate weighted comprehensive feature data. This embodiment inputs the weighted feature data into the lifespan prediction model along with nonlinear change characteristics and target time period operating data, and outputs the target data to obtain the lifespan prediction results.

[0054] As an example, the energy storage device life prediction method is introduced using a fully connected neural network model. The fully connected neural network model can include an embedding layer, an attention fusion layer, and an output layer. The embedding layer is used to map the target relevance index, target nonlinear characteristics, and operating data corresponding to the target energy storage device into a feature space of uniform dimensionality. The attention fusion layer is used to use the absolute value of the target relevance index as the attention weight and perform a weighted sum of the three types of features to obtain a fused feature. The output layer is used to map the fused feature to a life indicator and output the life prediction result.

[0055] Specifically, in actual application, the processing flow of the energy storage equipment life prediction model for the target correlation index, the target nonlinear change characteristics and the operating data corresponding to the target energy storage equipment within the target time period may include: the embedding layer maps the input data to a 128-dimensional unified feature space; the attention fusion layer uses the absolute value of the target correlation index as the weight to fuse the three types of features, capture the interactive dependency between the factors, and obtain the fused feature vector; the output layer may include two fully connected layers, wherein the first fully connected layer performs a linear transformation on the fused feature vector to obtain a 64-dimensional intermediate vector, and the second fully connected layer performs a linear transformation on the 64-dimensional intermediate vector to obtain the life prediction result.

[0056] From the above, it can be concluded that the embodiment of the present application can accurately screen out key factors that have a significant impact on the lifespan by calculating the correlation index between multiple factors and the lifespan of energy storage equipment, avoid redundant information interference, and lay a reliable foundation for prediction. The embodiment of the present application extracts nonlinear change characteristics based on the causal index, which can deeply capture the complex interactions and nonlinear relationships between multiple factors, and effectively make up for the shortcomings of traditional methods in nonlinear modeling. The embodiment of the present application inputs the correlation index, nonlinear change characteristics and operating data into the prediction model to achieve a more comprehensive and accurate characterization of the lifespan of energy storage equipment. Compared with traditional methods, it significantly improves the prediction accuracy, helps to optimize the operation and maintenance strategy of energy storage equipment, and ensures the stable and efficient operation of the energy storage system.

[0057] In one embodiment of the present application, the target relevance index is determined based on historical operating data of multiple energy storage devices in the following manner:

[0058] Extracting data corresponding to multiple factors based on historical operating data of multiple energy storage devices;

[0059] Obtain life indicator data for multiple energy storage devices, including battery health status, battery internal impedance, number of charge and discharge cycles, and battery capacity decay rate;

[0060] Align the lifespan indicator data with the data corresponding to multiple factors in the time dimension;

[0061] For each of the multiple factors, a first time period set is selected from a target historical time period based on the data change rate of the multiple factors; a correlation index between the factor and the equipment life is calculated based on the data corresponding to the factor in the first time period set and the life index data; the target historical time period is the data generation period corresponding to the historical operation data;

[0062] Based on the correlation index between each factor and the equipment life, a target correlation index is determined.

[0063] In this embodiment, data corresponding to the multiple factors are extracted based on the historical operation data of multiple energy storage devices, specifically including: for the historical operation data of each energy storage device, ambient temperature data, charge and discharge rate data, charge and discharge depth data, and usage time data are extracted.

[0064] This embodiment can calculate the battery health status based on parameters such as battery capacity and internal resistance through the built-in algorithm of the battery management system, regularly trigger full-capacity testing, and generate battery health status percentage data. This embodiment can use AC impedance spectroscopy technology to inject a small AC signal when the battery is in a static state, measure the voltage response at different frequencies, and calculate the internal resistance data. This embodiment can set a counter in the battery management system, and when the charge and discharge depth exceeds a specific threshold, it is counted as a complete cycle, and the number of cycles is accumulated and stored in real time. This embodiment can perform standard charge and discharge tests regularly, record the difference between the actual available capacity and the initial capacity, divide it by the test interval time, and obtain the capacity decay rate. This embodiment can synchronize the above data to the database according to the device number and timestamp to ensure time alignment with the operating factor data for subsequent analysis.

[0065] In this embodiment, for each of the multiple factors, a first time period set is selected from the target historical time period based on the data change rate of the multiple factors, specifically including:

[0066] For each of the multiple factors, a first data change rate is calculated based on data corresponding to the factor, and a second time period set is determined from the target historical time period based on the first data change rate; the second time period set includes time periods that meet a first condition; the first condition is that the first data change rate corresponding to a certain time period is greater than or equal to a first threshold;

[0067] The second data change rate in the second time period set is calculated based on the data corresponding to all factors except the factor in the multiple factors, and the first time period set is selected from the second time period set based on the second data change rate; the first time period set includes the time periods in the second time period set that meet the second condition; the second condition is that the second data change rate corresponding to a certain time period is less than or equal to the second threshold.

[0068] In this embodiment, the first data change rate calculates the degree of fluctuation of a single factor (such as ambient temperature) within each time period of historical operating data. Parameters may include indicators such as slope and standard deviation that characterize the magnitude of the change. The second time period set refers to the set of time periods where the first data change rate is greater than or equal to a first threshold. Parameters include the start and end times of each time period. The first threshold is a manually set critical value that determines whether the first data change rate is significant. The first threshold varies for different factors and must be determined based on experience or experimentation. The second data change rate is the weighted sum of the data change rates of all factors except the specified factor, resulting in a comprehensive change rate. Parameters may also be indicators that characterize the degree of change, such as the weighted standard deviation. The first time period set is selected from the second time period set. Time periods where the second data change rate is less than or equal to the second threshold are used for subsequent correlation index calculation. The second threshold is a manually set critical value used to screen the first time period set, determining whether the changes in other factors are relatively stable.

[0069] In this embodiment, the lifespan of energy storage devices is influenced by multiple factors. Excessive data changes in a single factor can interfere with the true impact of other factors. By setting a dual threshold to filter the first time period, first selecting periods where a particular factor changes significantly, and then selecting periods within these periods where other factors change relatively steadily, this can reduce the mutual interference of multiple factors, allowing the calculated correlation index to more accurately reflect the true relationship between that factor and lifespan.

[0070] For example, using ambient temperature as an example, this embodiment can calculate the first data rate of change of temperature within each month, such as the monthly temperature standard deviation. Setting a first threshold of 5°C (i.e., filtering out months with temperature fluctuations greater than or equal to 5°C), this results in a second time period set, for example, the high-temperature summer months. This embodiment then calculates the second data rate of change of other factors within the same time period, such as the weighted sum of the rate standard deviation and the depth fluctuation amplitude. Based on the set second threshold, this first time period is filtered out to include weeks from June to August with minimal rate and depth fluctuations, such as a weekly rate less than or equal to 1°C and a depth less than or equal to 70%. The correlation between temperature and battery health status is analyzed within this filtered first time period: the average temperature during this period is 35°C, and the battery health status drops from 92% to 91%. Combined with historical data under the same conditions, the correlation index between temperature and battery health status is calculated to be -0.8 (a strong negative correlation), indicating that high temperatures significantly accelerate battery degradation.

[0071] Similarly, repeat the above steps for the charge and discharge rate: select the time period when the rate change rate is greater than or equal to 0.5C / day and other factors are stable, and calculate its correlation index with the capacity decay rate to be 0.7 (strong positive correlation).

[0072] Similarly, perform corresponding steps for other factors.

[0073] In this embodiment, the life index data is data collected based on a first frequency, and the data corresponding to the multiple factors are data collected based on a second frequency; the second frequency is greater than the first frequency;

[0074] In this embodiment, aligning the life index data with the data corresponding to the multiple factors in the time dimension specifically includes: obtaining a forward time window;

[0075] For each collected life index data, performing a first operation;

[0076] The first operation includes:

[0077] Obtaining the collection timestamp of the life index data collected this time, and extracting the data within the forward time window from the data corresponding to the multiple factors;

[0078] Aggregate the data corresponding to multiple factors in the forward time window to obtain aggregated data sequences corresponding to the multiple factors;

[0079] Align the timestamps of the aggregated data series corresponding to multiple factors with the collection timestamps of the life indicator data.

[0080] In this embodiment, obtaining the forward time window may include: determining the forward time window based on a difference between collection timestamps of any two adjacent life indicator data.

[0081] In this embodiment, the first frequency refers to the frequency at which life indicator data is collected. The second frequency refers to the frequency at which operational factor data, such as ambient temperature and charge / discharge rate, are collected. The forward time window is a time range defined forward from each collection timestamp of life indicator data. The parameters include the window length, such as 1 day or 7 days. For example, if the collection frequency of life indicator data is once a month, then the difference between any two adjacent collection timestamps of life indicator data is one month, and the forward time window is one month.

[0082] Aggregated data series refers to the data series formed by statistically aggregating the data of each factor in the forward time window. The parameters may include statistics such as the mean, maximum value, minimum value, standard deviation, and other data characteristics corresponding to multiple factors.

[0083] Considering that life indicator data is collected infrequently, while operational factor data is collected infrequently, directly linking the two can easily lead to information mismatches. This embodiment uses a forward time window to define the time range, aggregates high-frequency operational factor data into a low-frequency data sequence, and aligns it with the collection timestamps of the low-frequency life indicator data. This not only preserves the comprehensive impact of operational factors but also accurately matches the two types of data in the time dimension, providing a reliable data foundation for subsequent correlation analysis and life prediction.

[0084] For example, in a large-scale lithium battery energy storage power station, life indicator data such as battery health status is collected once a week, while operating factor data such as ambient temperature and charge and discharge rate are collected once an hour.

[0085] This embodiment can set the forward time window length to 7 days, that is, each time the life indicator data is collected, the data is traced back 7 days. For example, if the battery health status data is collected on October 15, 2024, the data corresponding to October 8-14 will be extracted from the operating factor data. This embodiment can calculate the mean, maximum, and minimum values ​​of the temperature data during these 7 days; for the charge and discharge rate data, the daily average rate and the proportion of hours with high rate (greater than 1C) are calculated to obtain the aggregated data series for each factor.

[0086] In this example, the timestamp of the aggregated data series for each factor is set to October 14th to align it with the battery health status data collected at that time. Repeating the above steps completes the time dimension alignment of all life indicator data and operating factor data, providing complete and accurate data for subsequent calculation of the correlation index and construction of the life prediction model.

[0087] This embodiment uses dual thresholds to filter the first time period, reducing multi-factor interference and accurately calculating the correlation index. This embodiment also uses forward time windows to align high- and low-frequency data, preventing information mismatches. The combination of these two approaches effectively improves data quality and analysis accuracy, more accurately reflecting the relationship between various factors and energy storage device lifespan, thereby enhancing the accuracy of the lifespan prediction model and providing a reliable basis for equipment operation and maintenance.

[0088] In one embodiment of the present application, the process of calculating the causal index between each two factors among the multiple factors includes:

[0089] constructing a first autoregressive model based on the data corresponding to the first factor of the two factors;

[0090] constructing a second regression model based on the first autoregressive model, the data corresponding to the second factor of the two factors, and the optimal lag order;

[0091] Calculating the model residual sum of squares of the first autoregressive model, and calculating the model residual sum of squares of the second regression model;

[0092] The causal index between the two factors is determined based on the model residual sum of squares of the first autoregressive model and the model residual sum of squares of the second regression model.

[0093] In this embodiment, determining the causal index between the two factors based on the model residual sum of squares of the first autoregressive model and the model residual sum of squares of the second regression model includes:

[0094] Obtain the data sample size corresponding to the two factors;

[0095] Calculate the variance ratio statistic based on the model residual sum of squares of the first autoregressive model, the model residual sum of squares of the second regression model, the optimal lag order, and the data sample size of the data corresponding to the two factors;

[0096] The causal index between the two factors was determined based on the variance ratio statistic value.

[0097] In this embodiment, the first factor and the second factor are any two factors selected from multiple factors affecting the life of the energy storage device, such as ambient temperature and charge / discharge rate.

[0098] The first autoregressive model is constructed based on the historical data of the first factor and is used to describe the temporal changes of that factor. Parameters may include the model order and regression coefficient. The second regression model is constructed by adding the data of the second factor to the first autoregressive model. It is used to analyze the impact of the second factor on the first factor. Parameters may include the optimal lag order, which represents the time delay after which the second factor data affects the first factor. The model residual sum of squares is a measure of the error between the model's predicted value and the actual value. The smaller the model residual sum of squares, the better the model fit. The variance ratio statistic is calculated by combining the residual sum of squares of the first and second models, the optimal lag order, and the data sample size. It is used to quantify the strength of the causal relationship between the two factors.

[0099] In this embodiment, the lifespan of energy storage equipment is influenced by the interaction of multiple factors. By constructing a first autoregressive model and a second regression model, and comparing the change in model error before and after the addition of the second factor, this embodiment can determine whether the second factor significantly improves the predictive power of the first factor. If the error is significantly reduced after the addition of the second factor, it indicates that the second factor has a causal influence on the first factor. The variance ratio statistic and the causal index can quantify the extent of this influence, providing a basis for analyzing the causal relationship between factors.

[0100] For example, in a large lithium battery energy storage power station, to analyze the causal relationship between ambient temperature and charge and discharge rate, the following operations are performed:

[0101] This example acquires historical data for the past year, where ambient temperature is collected every hour and charge / discharge rate is collected every minute. The charge / discharge rate data is aggregated into hourly averages and aligned with the time frequency of the temperature data.

[0102] With ambient temperature as the first factor and charge / discharge rate as the second factor, we first attempted to construct multiple autoregressive models from the first to fifth order based on historical temperature data. By comparing the differences between the model predictions and the actual temperature values, we determined that the third-order autoregressive model had the smallest prediction error. This model was then used as the first autoregressive model, and the residual sum of squares of the first autoregressive model was calculated. This was done by squarely adding the difference between the predicted temperature and the actual temperature at each time point.

[0103] Determine the optimal lag order for the charge and discharge rate: Starting with a lag of 1 hour, gradually increasing the lag length, comparing the prediction results of the second regression model under different lag orders, and ultimately determining that the optimal lag order for the charge and discharge rate is 2 hours, meaning that the current temperature is affected by the charge and discharge rate 2 hours ago. Based on the third-order autoregressive model, add the charge and discharge rate with a 2-hour lag as an independent variable to construct a second regression model that includes the charge and discharge rate lag data, and calculate the sum of squared residuals of the second regression model.

[0104] This example compares the residual sums of squares of the two models. If the difference between the residual sums of squares of the second regression model and the first autoregressive model exceeds a threshold, it indicates that the charge and discharge rate has a causal effect on the ambient temperature. Combining the data sample size and the number of lags, this example further calculates a variance ratio statistic. If the variance ratio statistic is greater than a preset threshold, it indicates a strong causal relationship between the two. The causal index can be mapped to a specific value based on the size of the statistic, providing a basis for optimizing the power plant's temperature control and charge and discharge strategies.

[0105] Exemplarily, the calculation process of the variance ratio statistic value is: (the residual sum of squares of the first model - the residual sum of squares of the second model) ÷ the residual sum of squares of the second model × (data sample size - the number of model parameters).

[0106] This embodiment constructs autoregressive and regression models, compares error changes, and accurately determines the causal relationship between factors. It uses variance ratio statistics to quantify the impact intensity. This can effectively analyze the multi-factor interaction mechanism of energy storage equipment, avoid pseudo-correlation interference, and provide a scientific basis for power plants to formulate temperature control and charging and discharging strategies.

[0107] In one embodiment of the present application, the target nonlinear change characteristic is obtained by:

[0108] Constructing factor interaction terms based on the causal index and the multi-factor correlation vector, and splicing the factor interaction terms into the multi-factor correlation vector to obtain a spliced ​​vector;

[0109] Attention is allocated to each element in the concatenated vector based on the causal index to obtain the target vector;

[0110] The nonlinear variation characteristics between the target vector and the life of the energy storage device are extracted as the target nonlinear variation characteristics.

[0111] In this embodiment, the specific steps of extracting the nonlinear change characteristics between the multi-factor correlation vector and the life of the energy storage device based on the causal index can be implemented by an LSTM model based on the attention mechanism and feature intersection.

[0112] Exemplarily, the nonlinear variation characteristics between the multi-factor correlation vector and the life of the energy storage device are extracted based on the causal index, specifically including:

[0113] The causal index and multi-factor correlation vector are input into an LSTM model based on an attention mechanism and feature crossover to obtain the nonlinear variation characteristics between the multi-factor correlation vector and the device lifespan.

[0114] The LSTM model based on attention mechanism and feature cross includes input layer, feature cross layer, LSTM layer, attention layer, fully connected layer and output layer;

[0115] The feature cross layer is used to construct factor interaction terms based on the causal index, and the factor interaction terms are spliced ​​into the multi-factor association vector to obtain the target feature vector;

[0116] The LSTM layer is used to extract the temporal dependency features of the target feature vector at adjacent time steps to obtain the hidden state of each time step;

[0117] The attention layer is used to assign attention weights to the hidden states of each time step based on the causal index, and perform weighted summation of the hidden states of all time steps based on the attention weights to obtain the target vector.

[0118] In this embodiment, the factor interaction term is a new feature formed by combining factors with causal relationships based on the causal index. In specific implementation, all causal indices can be traversed to screen out all factors with causal relationships. For example, if the causal index of ambient temperature and charge and discharge rate meets the preset threshold condition, it is determined that if there is a causal relationship, the product of the two or other combinations can be used as the factor interaction term. If the ambient temperature and charge and discharge depth have a causal relationship, the product of the two or other combinations can also be used as the factor interaction term. The concatenated vector refers to the new vector formed after adding the factor interaction term to the multi-factor association vector. Attention allocation refers to assigning different weights to each element of the concatenated vector according to the size of the causal index to highlight important factors. The target vector is the vector obtained after attention allocation, which concentrates on the characteristics of factors that have a significant impact on the life of the equipment.

[0119] The LSTM model, based on the attention mechanism and feature crosstalk, is a deep learning model that integrates these two mechanisms. The input layer receives causal indices and multi-factor correlation vectors; the feature crosstalk layer constructs and concatenates factor interaction terms; the LSTM layer extracts temporal dependency features; the attention layer assigns weights; and the fully connected and output layers output nonlinear features.

[0120] In this example, the lifespan of energy storage devices is affected by the nonlinear interactions of multiple factors and time-series variations. This example constructs factor interaction terms using a causal index to capture these interactions. It also utilizes an attention mechanism to assign weights based on causal strength, highlighting key factors. The LSTM model excels at processing time-series data. Through collaboration across its layers, it can extract the complex nonlinear variations between multiple factors and device lifespan, providing more precise input for lifespan prediction and improving prediction accuracy.

[0121] Exemplarily, this embodiment inputs the calculated causal index of each factor and the constructed multi-factor association vector into the input layer of the LSTM model based on the attention mechanism and feature cross-talk, ensuring that the time sequence and dimension of the data meet the model requirements.

[0122] At the feature intersection layer, the model determines the causal relationship between factors based on the causal index. For pairs of factors with a causal relationship, the model generates factor interaction terms according to predefined rules. For example, this involves multiplying the data corresponding to the two factors or performing other combination operations. The model then concatenates the generated factor interaction terms into the original multi-factor association vector to obtain the target feature vector.

[0123] The target feature vector is input into the LSTM layer, and the model processes the data sequentially by time step. The LSTM layer uses internal memory cells and gating mechanisms to capture the temporal dependencies between adjacent time steps of the target feature vector, outputting a hidden state at each time step.

[0124] At the attention layer, the model assigns attention weights to the hidden states of each time step based on the causal index. Hidden states with stronger causal relationships receive higher weights. Based on these assigned weights, the model performs a weighted summation of the hidden states across all time steps to produce the final target vector.

[0125] The target vector is input into the fully connected layer for feature fusion and then outputted through the output layer to obtain the nonlinear change characteristics between the multi-factor correlation vector and the equipment life.

[0126] This embodiment constructs factor interaction terms through causal indices, accurately capturing nonlinear relationships between multiple factors. It also utilizes an attention mechanism to assign weights, highlighting key influencing factors. Furthermore, it incorporates LSTM to extract time series features, effectively handling dynamic data changes. This embodiment comprehensively explores the complex relationships between multiple factors and equipment lifespan, providing high-precision features for energy storage equipment lifespan prediction, improving prediction accuracy and reliability.

[0127] Corresponding to the energy storage device life prediction method in the above embodiment, Figure 2 This is a structural block diagram of the energy storage device life prediction system provided by an embodiment of the present application. For ease of explanation, only the parts related to the embodiment of the present application are shown. Figure 2 The energy storage device life prediction system 20 includes: an operation data acquisition module 21, a correlation index acquisition module 22, a nonlinear feature acquisition module 23 and a life prediction module 24.

[0128] The operation data acquisition module 21 is used to acquire the operation data corresponding to the target energy storage device within the target time period, where the target energy storage device is the energy storage device to be predicted;

[0129] A correlation index acquisition module 22 is configured to acquire a target correlation index, which is a correlation index between multiple factors and the life of the energy storage device. The multiple factors include ambient temperature, charge and discharge rate, charge and discharge depth, and usage time. The target correlation index is calculated based on historical operating data of multiple energy storage devices.

[0130] A nonlinear feature acquisition module 23 is configured to acquire a target nonlinear change feature, wherein the target nonlinear change feature is a nonlinear change feature between a multi-factor correlation vector extracted based on a causal index and the life of the energy storage device. The causal index is a causal index between every two factors in the multiple factors. The multi-factor correlation vector is constructed based on data corresponding to the multiple factors. The data corresponding to the multiple factors include ambient temperature data, charge and discharge rate data, charge and discharge depth data, and usage time data.

[0131] The life prediction module 24 is used to input the target correlation index, the target nonlinear change characteristics and the operating data corresponding to the target energy storage device in the target time period into the energy storage device life prediction model to perform device life prediction and obtain a life prediction result.

[0132] In one embodiment of the present application, the correlation index acquisition module 22 is specifically configured to extract data corresponding to multiple factors based on historical operation data of multiple energy storage devices;

[0133] Obtain life indicator data for multiple energy storage devices, including battery health status, battery internal impedance, number of charge and discharge cycles, and battery capacity decay rate;

[0134] Align the lifespan indicator data with the data corresponding to multiple factors in the time dimension;

[0135] For each of the multiple factors, a first time period set is selected from a target historical time period based on the data change rate of the multiple factors; a correlation index between the factor and the equipment life is calculated based on the data corresponding to the factor in the first time period set and the life index data; the target historical time period is the data generation period corresponding to the historical operation data;

[0136] Based on the correlation index between each factor and the equipment life, a target correlation index is determined.

[0137] In one embodiment of the present application, the correlation index acquisition module 22 is further configured to calculate, for each of the multiple factors, a first data change rate based on data corresponding to the factor, and determine a second time period set from the target historical time period based on the first data change rate; the second time period set includes time periods that meet a first condition; the first condition being that the first data change rate corresponding to a certain time period is greater than or equal to a first threshold;

[0138] The second data change rate in the second time period set is calculated based on the data corresponding to all factors except the factor in the multiple factors, and the first time period set is selected from the second time period set based on the second data change rate; the first time period set includes the time periods in the second time period set that meet the second condition; the second condition is that the second data change rate corresponding to a certain time period is less than or equal to the second threshold.

[0139] In one embodiment of the present application, the life index data is data collected based on a first frequency, and the data corresponding to the multiple factors are data collected based on a second frequency; the second frequency is greater than the first frequency; the correlation index acquisition module 22 is further configured to obtain a forward time window; for each collected life index data, a first operation is performed;

[0140] The first operation includes:

[0141] Obtaining the collection timestamp of the life index data collected this time, and extracting the data within the forward time window from the data corresponding to the multiple factors;

[0142] Aggregate the data corresponding to multiple factors in the forward time window to obtain aggregated data sequences corresponding to the multiple factors;

[0143] Align the timestamps of the aggregated data series corresponding to multiple factors with the collection timestamps of the life indicator data.

[0144] In one embodiment of the present application, the nonlinear feature acquisition module 23 is specifically configured to construct a first autoregressive model based on data corresponding to a first factor of the two factors;

[0145] constructing a second regression model based on the first autoregressive model, the data corresponding to the second factor of the two factors, and the optimal lag order;

[0146] Calculating the model residual sum of squares of the first autoregressive model, and calculating the model residual sum of squares of the second regression model;

[0147] The causal index between the two factors is determined based on the model residual sum of squares of the first autoregressive model and the model residual sum of squares of the second regression model.

[0148] In one embodiment of the present application, the nonlinear feature acquisition module 23 is further configured to acquire a data sample size of data corresponding to the two factors;

[0149] Calculate the variance ratio statistic based on the model residual sum of squares of the first autoregressive model, the model residual sum of squares of the second regression model, the optimal lag order, and the data sample size of the data corresponding to the two factors;

[0150] The causal index between the two factors was determined based on the variance ratio statistic value.

[0151] In one embodiment of the present application, the nonlinear feature acquisition module 23 is further configured to construct a factor interaction term based on the causal index and the multi-factor association vector, and to splice the factor interaction term into the multi-factor association vector to obtain a spliced ​​vector;

[0152] Attention is allocated to each element in the concatenated vector based on the causal index to obtain the target vector;

[0153] The nonlinear variation characteristics between the target vector and the life of the energy storage device are extracted as the target nonlinear variation characteristics.

[0154] See also Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided in one embodiment of the present application. Figure 3The electronic device 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of the modules in the above-mentioned system embodiments, such as Figure 2 The functions of the operating data acquisition module 21, the correlation index acquisition module 22, the nonlinear feature acquisition module 23 and the life prediction module 24 are shown.

[0155] It should be understood that in the embodiment of the present application, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0156] The input device 302 may include a touchpad, a fingerprint collection sensor (for collecting user fingerprint information and fingerprint direction information), a microphone, etc. The output device 303 may include a display (LCD, etc.), a speaker, etc.

[0157] The memory 304 may include a read-only memory and a random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store information about the type of energy storage device.

[0158] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiment of the present application can execute the implementation method described in the embodiment of the energy storage device life prediction method provided in the embodiment of the present application, and can also execute the implementation method of the electronic device 300 described in the embodiment of the present application, which will not be repeated here.

[0159] In another embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, all or part of the process of the method in the above embodiment is implemented. The computer program can also be used to instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of each of the above method embodiments are implemented. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium.

[0160] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the aforementioned embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the computer-readable storage medium can include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or is about to be output.

[0161] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0162] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0163] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the system embodiments described above are merely schematic. For example, the division of units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or can be an electrical, mechanical or other form of connection.

[0164] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0165] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0166] The above are only specific embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for predicting the life of an energy storage device, characterized in that: include: Obtaining operating data corresponding to a target energy storage device within a target time period, wherein the target energy storage device is the energy storage device to be predicted; Obtaining a target correlation index, where the target correlation index is a correlation index between multiple factors and the life of the energy storage device, the multiple factors including ambient temperature, charge and discharge rate, charge and discharge depth, and usage time, and the target correlation index is calculated based on historical operating data of the multiple energy storage devices; Obtaining a target nonlinear change characteristic, wherein the target nonlinear change characteristic is a nonlinear change characteristic between a multi-factor correlation vector extracted based on a causal index and the life of the energy storage device, wherein the causal index is a causal index between every two factors among the multiple factors, and the multi-factor correlation vector is constructed based on data corresponding to the multiple factors; the data corresponding to the multiple factors include ambient temperature data, charge and discharge rate data, charge and discharge depth data, and usage time data; Inputting the target correlation index, the target nonlinear change characteristic, and the operating data corresponding to the target energy storage device within the target time period into an energy storage device life prediction model to perform device life prediction and obtain a life prediction result; The target relevance index is determined based on historical operating data of multiple energy storage devices in the following manner: Extracting data corresponding to the multiple factors based on historical operation data of multiple energy storage devices; Acquire life index data of the plurality of energy storage devices, the life index data including battery health status, battery internal impedance, number of charge and discharge cycles, and battery capacity attenuation rate of the plurality of energy storage devices; Aligning the life index data with the data corresponding to the multiple factors respectively in the time dimension; For each of the multiple factors, a first time period set is selected from a target historical time period based on the data change rate of the multiple factors; a correlation index between the factor and the equipment life is calculated based on the data corresponding to the factor in the first time period set and the life index data; the target historical time period is a data generation period corresponding to the historical operation data; Determining the target correlation index based on the correlation index between each factor and the equipment life; For each of the multiple factors, selecting a first time period set from a target historical time period based on the data change rate of the multiple factors includes: For each of the multiple factors, a first data change rate is calculated based on data corresponding to the factor, and a second time period set is determined from the target historical time period based on the first data change rate; the second time period set includes time periods that meet a first condition; the first condition is that the first data change rate corresponding to a certain time period is greater than or equal to a first threshold; Based on the data corresponding to all factors except the factor in the multiple factors, the second data change rate in the second time period set is calculated, and based on the second data change rate, the first time period set is selected from the second time period set; the first time period set includes the time periods in the second time period set that meet the second condition; the second condition is that the second data change rate corresponding to a certain time period is less than or equal to the second threshold.

2. The energy storage device life prediction method according to claim 1, characterized in that: The life index data is data collected based on a first frequency, and the data corresponding to the multiple factors are data collected based on a second frequency; The second frequency is greater than the first frequency; The aligning of the life index data with the data corresponding to the multiple factors in the time dimension includes: Get the forward time window; For each collected life index data, performing a first operation; The first operation includes: Obtaining a collection timestamp of the life index data collected this time, and extracting data within the forward time window from the data corresponding to the multiple factors respectively; Aggregating the data corresponding to the multiple factors in the forward time window to obtain aggregated data sequences corresponding to the multiple factors; The timestamps of the aggregated data sequences corresponding to the multiple factors are aligned with the collection timestamps of the life index data.

3. The energy storage device life prediction method according to claim 1, characterized in that: The calculation process of the causal index between each two factors in the plurality of factors includes: constructing a first autoregressive model based on the data corresponding to the first factor of the two factors; constructing a second regression model based on the first autoregressive model, the data corresponding to the second factor of the two factors, and the optimal lag order; Calculating the model residual sum of squares of the first autoregressive model, and calculating the model residual sum of squares of the second regression model; A causal index between the two factors is determined based on the model residual sum of squares of the first autoregressive model and the model residual sum of squares of the second regression model.

4. The energy storage device life prediction method according to claim 3, characterized in that: The determining the causal index between the two factors based on the model residual sum of squares of the first autoregressive model and the model residual sum of squares of the second regression model includes: Obtain the data sample size corresponding to the two factors; Calculating a variance ratio statistic based on the model residual sum of squares of the first autoregressive model, the model residual sum of squares of the second regression model, the optimal lag order, and the data sample size of the data corresponding to the two factors; A causal index between the two factors is determined based on the variance ratio statistic value.

5. The energy storage device life prediction method according to claim 1, characterized in that: The target nonlinear change characteristics are obtained in the following way: constructing a factor interaction term based on the causal index and the multi-factor association vector, and splicing the factor interaction term into the multi-factor association vector to obtain a spliced ​​vector; Allocating attention to each element in the concatenated vector based on the causal index to obtain a target vector; A nonlinear variation feature between the target vector and the life of the energy storage device is extracted as the target nonlinear variation feature.

6. A life prediction system for energy storage equipment, characterized in that: include: An operation data acquisition module is used to acquire operation data corresponding to a target energy storage device within a target time period, wherein the target energy storage device is the energy storage device to be predicted; a correlation index acquisition module, configured to acquire a target correlation index, wherein the target correlation index is a correlation index between multiple factors and the life of the energy storage device, the multiple factors including ambient temperature, charge and discharge rate, charge and discharge depth, and usage time, and the target correlation index is calculated based on historical operating data of multiple energy storage devices; A correlation index acquisition module, specifically configured to extract data corresponding to the multiple factors based on historical operation data of multiple energy storage devices; Acquire life index data of the plurality of energy storage devices, the life index data including battery health status, battery internal impedance, number of charge and discharge cycles, and battery capacity attenuation rate of the plurality of energy storage devices; Aligning the life index data with the data corresponding to the multiple factors respectively in the time dimension; For each of the multiple factors, a first time period set is selected from a target historical time period based on the data change rate of the multiple factors; a correlation index between the factor and the equipment life is calculated based on the data corresponding to the factor in the first time period set and the life index data; the target historical time period is a data generation period corresponding to the historical operation data; Determining the target correlation index based on the correlation index between each factor and the equipment life; The correlation index acquisition module is further configured to calculate, for each of the multiple factors, a first data change rate based on data corresponding to the factor, and determine a second time period set from the target historical time period based on the first data change rate; the second time period set includes time periods that meet the first condition; The first condition is that the first data change rate corresponding to a certain period is greater than or equal to a first threshold; calculating a second data change rate in a second time period set based on data corresponding to all factors except the factor in the multiple factors, and selecting a first time period set from the second time period set based on the second data change rate; the first time period set includes time periods in the second time period set that meet a second condition; the second condition being that the second data change rate corresponding to a certain time period is less than or equal to a second threshold; a nonlinear feature acquisition module for acquiring a target nonlinear change feature, wherein the target nonlinear change feature is a nonlinear change feature between a multi-factor association vector extracted based on a causal index and the life of the energy storage device, wherein the causal index is a causal index between every two factors among the multiple factors, and the multi-factor association vector is constructed based on data corresponding to the multiple factors; the data corresponding to the multiple factors include ambient temperature data, charge and discharge rate data, charge and discharge depth data, and usage time data; The life prediction module is used to input the target correlation index, the target nonlinear change characteristics and the operating data corresponding to the target energy storage device in the target time period into the energy storage device life prediction model to perform equipment life prediction and obtain a life prediction result.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images