Battery Performance Degradation Test Method, Device and Equipment for Energy Storage Batteries

Abstract

The present invention relates to the technical field of battery performance testing, and discloses a battery performance attenuation testing method, device and equipment for an energy storage battery. The method includes: constructing a battery performance measurement data matrix during the charge and discharge process of the energy storage battery, and calculating the optimal value of the noise variance through a zero-mean censored normal distribution model to obtain a battery performance noise feature matrix; performing mutual information calculation of performance parameters to obtain a battery performance key feature set, and constructing a battery attenuation feature vector of operating condition parameters and performance attenuation rate; extracting the weighted adaptive distances of time interval, operating condition difference and performance difference according to the battery attenuation feature vector, and creating a battery attenuation feature association network; classifying the performance attenuation mode based on the battery attenuation feature association network and the battery attenuation feature vector to obtain the classification result of the performance attenuation mode of the energy storage battery. The present invention comprehensively extracts the battery performance attenuation features and realizes the accurate classification of the performance attenuation mode.

Description

Technical Field

[0001] The present invention relates to the technical field of battery performance testing, and particularly to a method, device, and equipment for testing the performance degradation of energy storage batteries. Background Art

[0002] With the rapid development of the new energy industry, energy storage batteries are increasingly widely used in power systems, and the accurate evaluation of their performance degradation characteristics is of great significance for ensuring the safe operation of energy storage systems. At present, energy storage batteries face complex and changeable working conditions during actual operation, resulting in non-linear and time-varying characteristics of battery performance degradation, which makes it difficult for traditional performance degradation evaluation methods to accurately capture the dynamic process of battery performance degradation.

[0003] During the process of testing the performance degradation of energy storage batteries, the measured data often contains a large amount of noise information, which comes from multiple aspects such as the accuracy error of measurement equipment, environmental interference, and working condition fluctuations. At the same time, due to the complex coupling relationship between the performance parameters of energy storage batteries, it is difficult to extract effective performance degradation characteristics from the original test data, which seriously affects the accuracy of judging the performance degradation trend. Summary of the Invention

[0004] The present invention provides a method, device, and equipment for testing the performance degradation of energy storage batteries. The present invention comprehensively extracts the performance degradation characteristics of the battery and realizes the accurate classification of the performance degradation mode.

[0005] In a first aspect, the present invention provides a method for testing the performance degradation of an energy storage battery, and the method for testing the performance degradation of the energy storage battery includes:

[0006] Construct a battery performance measurement data matrix during the charge and discharge process of the energy storage battery, and calculate the optimal value of the noise variance through a zero-mean censored normal distribution model to obtain a battery performance noise feature matrix;

[0007] Based on the battery performance noise feature matrix, calculate the mutual information of performance parameters to obtain a key battery performance feature set, and construct a battery decay feature vector of working condition parameters and performance decay rate;

[0008] Extract the weighted adaptive distances of time interval, working condition difference, and performance difference according to the battery decay feature vector, and create a battery decay feature association network;

[0009] Based on the battery decay feature association network and the battery decay feature vector, perform performance degradation mode classification to obtain a classification result of the performance degradation mode of the energy storage battery.

[0010] In a second aspect, the present invention provides a battery performance degradation test device for an energy storage battery, and the battery performance degradation test device for the energy storage battery includes:

[0011] A construction module, configured to construct a battery performance measurement data matrix during the charge and discharge process of the energy storage battery, and calculate an optimal value of the noise variance through a zero-mean censored normal distribution model to obtain a battery performance noise feature matrix;

[0012] A calculation module, configured to calculate the mutual information of performance parameters based on the battery performance noise feature matrix to obtain a battery performance key feature set, and construct a battery degradation feature vector of the operating condition parameters and the performance degradation rate;

[0013] A creation module, configured to extract a weighted adaptive distance of the time interval, the operating condition difference, and the performance difference according to the battery degradation feature vector, and create a battery degradation feature association network;

[0014] A classification module, configured to classify the performance degradation mode based on the battery degradation feature association network and the battery degradation feature vector to obtain a classification result of the performance degradation mode of the energy storage battery.

[0015] In a third aspect of the present invention, a computer device is provided, including: a memory and at least one processor, and instructions are stored in the memory; the at least one processor calls the instructions in the memory so that the computer device executes the above-mentioned battery performance degradation test method for the energy storage battery.

[0016] In the technical solution provided by the present invention, the present invention models the measurement noise by adopting a zero-mean censored normal distribution model, and combines the maximum entropy principle to calculate the optimal value of the noise variance, realizing the effective elimination of noise in the measurement data. At the same time, Butterworth low-pass filtering and outlier processing are adopted to ensure the data quality; through a feature dimensionality reduction method based on the mutual information theory and the sequential forward selection algorithm, and an integrated feature metric function of the operating condition parameters and the performance degradation rate, the battery performance degradation features are comprehensively extracted; the weighted adaptive distance calculation and the community structure adjacency matrix construction method are introduced, combined with the Louvain community discovery algorithm, to reveal the internal connection between the performance degradation features; a multi-scale network feature extraction method is adopted to describe the performance degradation features from three levels of global, node, and community, and an accurate classification of the performance degradation mode is realized through a radial basis kernel function support vector machine classifier. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a schematic diagram of the steps of the battery performance degradation test method for the energy storage battery in the embodiment of the present invention;

[0019] Figure 2 It is a schematic diagram of the structure of the battery performance degradation test device for the energy storage battery in the embodiment of the present invention;

[0020] Figure 3 It is a schematic block diagram of the structure of the computer device in the embodiment of the present invention. Specific embodiments

[0021] The embodiments of the present invention provide a battery performance degradation test method, device and equipment for an energy storage battery. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims and above-mentioned accompanying drawings of the present invention are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order other than those illustrated or described here. In addition, the terms "include" or "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or equipment that includes a series of steps or units does not necessarily have to be limited to those clearly listed steps or units, but may include other steps or units not clearly listed or inherent to these processes, methods, products or equipment.

[0022] For ease of understanding, the following describes the specific process of the embodiments of the present invention. Please refer to Figure 1 , an embodiment of the battery performance degradation test method for the energy storage battery in the embodiment of the present invention includes:

[0023] Step S1: Construct a battery performance measurement data matrix during the charge and discharge process of the energy storage battery, and calculate the optimal value of the noise variance through the zero-mean censored normal distribution model to obtain the battery performance noise feature matrix;

[0024] It can be understood that the execution subject of the present invention can be a battery performance degradation test device for an energy storage battery, or a terminal or a server. Specifically, it is not limited here. The embodiments of the present invention will be described by taking the server as the execution subject as an example.

[0025] Specifically, data such as voltage, current, temperature, internal resistance value, and state of charge of the energy storage battery are collected during the charging and discharging process, and these data are recorded at certain time intervals to form an original data sequence. The original data sequence is segmented, and the entire data set is divided into multiple subsequences along the time axis to form a time series data set. Signal processing is performed on the time series data set to filter out high-frequency noise and random interference, and improve the smoothness and stability of the data. Signal processing methods include wavelet transform, median filtering, Kalman filtering, etc. Among them, wavelet transform effectively removes burst noise, while Kalman filtering is suitable for processing random noise in dynamic systems. During the filtering process, appropriate filtering windows and parameters are selected according to the data characteristics to ensure that the noise suppression effect does not cause loss of useful signals, and a filtered data set is obtained. Outliers in the filtered data set are identified and corrected. Outliers are manifested as mutation points, drift points, or outliers. Statistical methods such as Z-score detection, box plot analysis, etc., or combined with machine learning methods such as isolation forest or LOF (Local Outlier Factor) algorithm are used to more accurately judge outliers. Appropriate correction strategies are adopted, such as interpolating using data at adjacent times or performing regression prediction based on historical trends to fill in the abnormal data points, and a corrected data set is obtained. All performance parameters in the corrected data set are normalized to eliminate the dimension difference, so that different features have the same scale in subsequent calculations. Through normalization, it is avoided that some parameters have too much influence on the analysis results due to large numerical ranges, thereby improving the stability and calculation efficiency of the model, and a normalized data set is obtained. The normalized data set is matrix-reconstructed with the sampling time points as row vectors and performance parameters as column vectors to obtain a battery performance measurement data matrix, which reflects the trend of battery performance changing with time. A zero-mean censored normal distribution model is constructed based on the battery performance measurement data matrix to characterize the noise characteristics of the battery performance data. The zero-mean censored normal distribution model is a statistical model suitable for processing noisy data. It assumes that the noise in the data follows a normal distribution, and the data mean is made zero by censoring the abnormal points, improving the accuracy of noise estimation. During the process of constructing the model, the mean and variance of the battery performance measurement data matrix are calculated, and extreme values are removed by censoring the abnormal data points to make the data more in line with the assumption of normal distribution. On this basis, the optimal value of the noise variance is calculated using the principle of maximum entropy. Among all possible probability distributions, the distribution that satisfies the known constraints and has the maximum entropy is selected to ensure that the model makes the smallest assumptions about the unknown information. During the calculation process, the sample entropy is calculated based on the battery performance measurement data matrix, and the maximum entropy distribution is solved under the constraint conditions to obtain the optimal variance estimate value of the noise. The battery performance noise characteristic matrix is obtained through this method.

[0026] The battery performance measurement data matrix is processed by data chunking, which is divided according to a fixed time interval or based on changes in the battery operating state, so that each data chunk represents a relatively stable operating condition. Each data chunk is centered, that is, the mean of each performance parameter in the data chunk is calculated, and the original data is subtracted by the corresponding mean, so that the mean of the processed data chunk is zero, obtaining a zero-mean data chunk. The censored interval is divided for the zero-mean data chunk to obtain the censored interval. Statistical analysis methods are used, such as setting the censored range based on the standard deviation threshold, or using an adaptive method to dynamically determine the censored interval according to the internal change trend of the data chunk. Through the censored interval division, extremely deviated noise points are effectively removed while maintaining the representativeness of the data. Based on the censored interval division, a zero-mean censored normal distribution model is constructed. The basic assumption of this model is that the true distribution of the data is approximated by a normal distribution, and the noise part is adjusted through the censoring operation, so that the model can more accurately reflect the actual statistical characteristics of the battery performance data. The maximum entropy principle is used to construct and numerically optimize the objective function for optimizing the noise variance of the zero-mean censored normal distribution model to obtain the optimal value of the noise variance. The maximum entropy principle is to select the distribution with the highest entropy value as the optimal distribution among all possible probability distributions, because the maximization of entropy means minimizing the assumption of unknown information, thus avoiding artificial bias. In the specific implementation process, the entropy value of the battery performance measurement data matrix is calculated, and constraint conditions are established, such as the mean and variance ranges of the data, and then the probability distribution that maximizes the entropy value is solved under these constraint conditions. Through numerical optimization methods, such as the Lagrange multiplier method or the gradient descent algorithm, the optimal value of the noise variance is efficiently calculated and used for noise analysis. A white noise channel model is constructed according to the zero-mean censored normal distribution model and the optimal value of the noise variance. The white noise channel model can simulate the noise propagation process of the battery performance data and estimate the influence of different noise sources on the data. By constructing this model, the systematic changes and random noise components in the battery performance data are effectively separated, so as to more accurately describe the evolution trend of the battery performance over time. After the construction of the white noise channel model is completed, the noise distribution characteristics are calculated to extract the noise entropy values of each performance parameter. The noise entropy value measures the degree to which different performance parameters are affected by noise. The higher the entropy value, the greater the random fluctuation of the parameter and the higher the possibility of being interfered by noise. To calculate the noise entropy value, the information entropy of the noise distribution characteristics is calculated based on the Shannon entropy formula to obtain the noise entropy value of each performance parameter. The noise entropy values are constructed into a noise entropy feature set, and the noise entropy feature set is associated and mapped with the corresponding performance parameters to establish the relationship between the noise characteristics and the battery performance. On this basis, matrix transformation operations are performed to synthesize the noise entropy characteristics of multiple performance parameters, and finally the battery performance noise feature matrix is obtained.

[0027] Step S2: Calculate the mutual information of performance parameters based on the battery performance noise feature matrix to obtain the key feature set of battery performance, and construct the battery decay feature vector of operating conditions parameters and performance decay rate;

[0028] Specifically, calculate the marginal entropy for each performance parameter in the battery performance noise feature matrix. Marginal entropy is an important indicator for measuring the information content of a single parameter and reflects the distribution uncertainty of that parameter itself. Construct a joint entropy based on the marginal entropy to measure the coupling relationship between different parameters, forming a set of performance parameter entropy values. The calculation of joint entropy is based on the joint probability distribution of multiple parameters, ensuring that the dependencies between multiple parameters are considered. Construct a mutual information calculation formula based on the set of performance parameter entropy values. Mutual information measures the degree of information sharing between two performance parameters, that is, whether the information content of one parameter can be predicted by another parameter. The calculation of mutual information utilizes marginal entropy and joint entropy. Specifically, by calculating the joint entropy of each pair of performance parameters and combining the marginal entropy of individual parameters, a mutual information matrix is obtained. Each element of the mutual information matrix represents the magnitude of the mutual information between two parameters. Optimize the mutual information matrix by subtracting the corresponding noise entropy value to eliminate the interference of noise on the information relationship between parameters, obtaining the net information feature of performance parameters. Based on the net information feature of performance parameters, calculate the correlation metric and redundancy to analyze the information relationship between different performance parameters. The correlation metric is used to measure the dependence between parameters, while redundancy is used to eliminate parameters with high information overlap to ensure that the finally selected features are the most representative. Therefore, construct a feature evaluation function that comprehensively considers the net information feature, correlation metric, and redundancy to evaluate the importance of each performance parameter in battery performance degradation. Execute the sequential forward selection algorithm on the feature evaluation function. In each iteration, select the feature that maximizes the F-value increment and continuously expand the feature set. The F-value increment refers to the contribution of the newly added feature to the overall information content. The larger the increment, the more obvious the role of the feature in describing the battery performance degradation pattern. In each iteration, calculate the F-value and mark the features with F-values greater than the first target value as candidate features, forming a feature candidate set. The feature candidate set contains a group of performance parameters that may be the most representative. Conduct a correlation analysis on the feature candidate set to screen out the most representative feature combination. Calculate the correlation coefficients between candidate features and select the feature combination with the absolute value of the correlation coefficient greater than the second target value to ensure that the finally selected features are complementary in terms of information content rather than highly correlated redundant features. Through this process, a set of key features of battery performance is obtained. Based on the set of key features of battery performance, construct an integrated feature metric function for operating condition parameters and performance degradation rate to comprehensively measure the influence of different features on battery degradation. The construction of the integrated feature metric function adopts a weighted method, where the weight assignment is based on the importance measure of features, such that key features occupy a larger weight ratio in the calculation process. After constructing the integrated feature metric function, process the local extreme points of the function through density clustering to identify the key patterns of battery degradation features. The density clustering method identifies the clustering centers based on the distribution density of data, thereby more effectively extracting the representative patterns of battery degradation features.Through this step, a battery degradation feature vector is obtained, which describes the degradation characteristics of the battery under different operating conditions.

[0029] Extract operating condition parameters from the key feature set of battery performance and calculate the corresponding performance degradation rate. The operating condition parameters include the charge-discharge rate, temperature change, SOC (state of charge) range, etc. of the battery, while the performance degradation rate involves key indicators such as the capacity retention rate and the internal resistance growth rate that characterize the battery health state. After extracting these parameters, set a balance factor to establish a trade-off relationship between the operating condition parameters and the performance degradation rate, ensuring that the constructed integrated feature metric function comprehensively reflects the operating state and degradation characteristics of the battery. The selection of the balance factor is based on the importance of different parameters to the battery life, and all features are integrated into a metric function through a weighted method to form a unified evaluation standard. Set a sliding time window for the integrated feature metric function to dynamically capture the time characteristics of battery performance degradation. The setting of the sliding time window helps to perform a localized analysis of the data to more precisely identify the performance change trend. Within each time window, calculate the local mean and standard deviation to obtain the fluctuation characteristics of the data on a short time scale. The local mean reflects the central value of the features during this time period, while the standard deviation is used to measure the degree of data fluctuation. Through these local statistical features, determine the trend of battery performance evolution over time. Based on the local statistical features, set local threshold conditions to screen out the key time points representing the battery degradation mode. Set a threshold range such that the time points where the local mean and standard deviation change exceed a certain range are marked as candidate feature points. These candidate feature points correspond to the critical moments when the battery performance changes during operation. Through this process, initially screen out the most representative time points during the battery decay process. Perform a density clustering algorithm on the candidate feature point set with a preset time span as the clustering radius to identify the patterns existing during the battery decay process. Based on the density distribution of the data points, divide the high-density regions into clusters and filter out the isolated points to ensure that the obtained feature points have high representativeness. To perform density clustering, preset a time span as the clustering radius and perform cluster division based on the point density and distance threshold. If the number of candidate feature points around a certain time point exceeds a certain threshold, then classify this point as a core point and assign the nearby candidate points to the same cluster. Through this step, aggregate the feature points within a similar time range together to form a series of stable decay patterns. Based on the feature point clustering results, select the center point of each cluster and record the time stamps, operating condition parameters, and performance parameters corresponding to these center points. The selection of the center point uses the point with the highest density within the cluster, or selects the point in the cluster that best represents the overall trend to ensure that the selected feature points accurately reflect the battery decay process. Each center point corresponds to a critical moment of battery performance change and is associated with specific operating conditions and performance parameters to form a sequence of key feature points. Reconstruct these key feature points in chronological order and represent them in vector form to obtain the battery decay feature vector.

[0030] Step S3: Extract the weighted adaptive distances of time interval, operating condition difference, and performance difference according to the battery degradation feature vector, and create a battery degradation feature association network;

[0031] Specifically, identify adjacent feature points in the battery degradation feature vector, and calculate the time interval, operating condition difference, and performance difference between them. The time interval reflects the time characteristics of battery degradation, the operating condition difference describes the changes of the battery under different operating conditions, and the performance difference measures the fluctuations in the health state of the battery at different time points. Weight these three differences so that the contributions of different factors are dynamically adjusted according to the specific usage scenario of the battery. The calculation method of the weighted adaptive distance uses the normalized Euclidean distance or Mahalanobis distance to ensure that various factors are comparable on the same numerical scale, and form a feature point distance matrix. Each element of this matrix represents the weighted distance between any two feature points. Set a connection threshold for the feature point distance matrix to determine which feature points are close enough in distance to establish an association. The connection threshold is set through the statistical characteristics of the data or an empirical threshold, so that most feature points with similar degradation patterns form connections. For all feature point pairs with a distance less than this threshold, mark them as connection point pairs to form an initial connection relationship set. Based on the initial connection relationship set, construct an adjacency matrix A, where the matrix element A ijIt takes the value of 1 when the feature points i and j are connected, and 0 otherwise, forming a binary adjacency matrix. During the process of constructing the adjacency matrix, sparsity control is carried out to ensure that the network structure has a certain connectivity and avoid affecting the accuracy of community division due to too many isolated nodes. Calculate the degree centrality and betweenness centrality of each node in the binary adjacency matrix. Degree centrality represents the number of neighbors directly connected to a node and is an important indicator to measure the direct influence of a node in the network, while betweenness centrality describes the degree to which the node acts as a bridge in the network, that is, how many shortest paths it plays a connecting role in. Normalize and weight-sum the calculated degree centrality and betweenness centrality to obtain the importance index of the node, which measures the core degree of different feature points in the whole network. Execute the Louvain community discovery algorithm on the binary adjacency matrix based on the node importance index, and perform community division through the optimization of the modularity Q value. The Louvain algorithm is an efficient modular optimization algorithm that automatically detects groups of nodes with high-density connections, that is, communities, based on the network topology. The modularity Q value is a key indicator to measure the quality of community division. The higher its value, the higher the connection density within the divided communities, and the sparser the connections between different communities. When executing the Louvain algorithm, continuously optimize the modularity Q value to obtain the optimal community division result, so as to ensure that the battery feature points of different decay modes can belong to different communities. Combine the topological features of the network community structure with the binary adjacency matrix to construct a battery decay feature correlation network. The network community structure provides clustering information of feature points, while the adjacency matrix describes the connection relationship between points. The combination of the two effectively characterizes the correlation pattern of battery decay features in time series and space.

[0032] Step S4: Based on the battery decay feature correlation network and the battery decay feature vector, perform performance decay mode classification to obtain the classification result of the energy storage battery performance decay mode.

[0033] Specifically, global topological features are extracted from the battery degradation feature correlation network, and a network global structure feature vector is constructed based on this topological information. The global topological features include the average shortest path length, network diameter, network density, and global clustering coefficient. Among them, the average shortest path length measures the average distance between nodes in the network, reflecting the connectivity of the network. The network diameter represents the maximum shortest path length, characterizing the overall expansion range of the network. The network density is used to describe the ratio between the actual existing edges and the maximum number of possible edges in the network, reflecting the tightness of the network. The global clustering coefficient measures the proportion of triangular relationships between nodes to describe the overall aggregation of the network. By calculating these global topological features and integrating them into the network global structure feature vector, the macroscopic structural characteristics of the battery degradation feature correlation network are comprehensively characterized. Calculate the node-level topological features of the battery degradation feature correlation network and splice them with the global structure feature vector to construct a node feature matrix. The node-level topological features include the node degree distribution, betweenness centrality, and eigenvector centrality. Among them, the node degree distribution describes the distribution of the connection numbers of each node, reflecting the differences in the importance of nodes in the network. The betweenness centrality is used to measure the bridging role of nodes in the shortest path, characterizing the control ability of nodes in the network. The eigenvector centrality is measured based on the centrality of neighbor nodes, measuring the influence of nodes in the entire network. By calculating these node-level topological features and splicing them into the same feature matrix with the global topological features, it is ensured that the model considers both the overall network structure and the characteristics of local nodes when performing degradation mode classification. Extract the community-level topological features of the battery degradation feature correlation network and perform feature combination with the node feature matrix to obtain a multi-scale network feature vector. The community-level topological features mainly include the number of communities, community size distribution, and connection strength between communities. Among them, the number of communities represents the number of communities identified in the network, reflecting the diversity of degradation modes. The community size distribution is used to describe the number of nodes included in different communities, thereby revealing the scale characteristics of degradation modes. The connection strength between communities measures the degree of association between different communities, reflecting the transition characteristics of the battery between different degradation modes. By combining these community-level topological features with the node feature matrix, a multi-scale network feature vector is formed to more comprehensively describe the structural characteristics of the battery degradation mode. Input the multi-scale network feature vector and the battery degradation feature vector into the input layer of the radial basis kernel function support vector machine classifier to form a standardized input feature. The radial basis kernel function support vector machine realizes the distinction of different degradation modes by constructing the optimal classification hyperplane in the high-dimensional feature space. When inputting features, all features are normalized to ensure that features with different dimensions are compared on the same scale and to avoid excessive influence of features with too large numerical ranges on model training. The radial basis kernel function is used to perform a non-linear mapping on the standardized input features, transforming the original features into a high-dimensional space to improve the separability of classification.After the kernel function mapping of the features is completed, based on the obtained kernel function mapping features, a classification decision function needs to be executed to classify the performance degradation modes of energy storage batteries. The classification decision function optimizes the classification hyperplane of the support vector machine, combines the distance information between different modes, and maximizes the interval between categories to improve the robustness of classification. Through this classification model, different performance degradation modes of energy storage batteries are accurately identified and the corresponding classification results are output.

[0034] In the embodiments of the present invention, the present invention models the measurement noise by using a zero-mean censored normal distribution model, calculates the optimal value of the noise variance in combination with the maximum entropy principle, realizes the effective elimination of noise in the measurement data, and at the same time uses Butterworth low-pass filtering and outlier processing to ensure the data quality; through a feature dimensionality reduction method based on the mutual information theory and the sequential forward selection algorithm, and an integrated feature metric function of operating condition parameters and performance degradation rate, the battery performance degradation features are comprehensively extracted; a weighted adaptive distance calculation and community structure adjacency matrix construction method are introduced, combined with the Louvain community discovery algorithm, to reveal the internal relationship between the performance degradation features; a multi-scale network feature extraction method is used to describe the performance degradation features from three levels: global, node and community, and an accurate classification of the performance degradation mode is realized through a radial basis kernel function support vector machine classifier.

[0035] In a specific embodiment, the process of executing step S1 may specifically include the following steps:

[0036] Collect data on the voltage, current, temperature, internal resistance value and state of charge of the energy storage battery during the charge and discharge process to obtain an original data sequence;

[0037] Segment the original data sequence to obtain a time series data set, and perform signal processing on the time series data set to obtain a filtered data set;

[0038] Identify and correct the outliers in the filtered data set to obtain a corrected data set, and perform normalization processing on all performance parameters in the corrected data set to obtain a normalized data set;

[0039] Reconstruct the normalized data set into a matrix with the sampling time points as row vectors and the performance parameters as column vectors to obtain a battery performance measurement data matrix;

[0040] Construct a zero-mean censored normal distribution model based on the battery performance measurement data matrix, and calculate the optimal value of the noise variance through the maximum entropy principle to obtain a battery performance noise feature matrix.

[0041] Specifically, the voltage is collected in real time during the operation of the battery , current , temperature , internal resistance and state of charge as basic performance parameters. Assume that at time , the observed data at each sampling point is expressed as:

[0042]

[0043] Then the original data sequence within the entire test period is expressed as:

[0044]

[0045] where represents the total number of sampling points. Segment the original data sequence. Based on a fixed time window or the start and end points of charge and discharge cycles for partitioning. For example, if a fixed time window is adopted, the resulting time-series data set after partitioning is expressed as:

[0046]

[0047] where represents the data set of the th time period, is the number of sampling points included in each segment. Perform signal processing on the time-series data set to filter out high-frequency interference components. Use low-pass filtering to effectively smooth the data. For example, perform moving average filtering on voltage data using the following formula:

[0048]

[0049] where is the smoothed voltage data, is the window width, which determines the smoothing degree of the filtering. Similarly, perform corresponding filtering processing on other parameters such as current , temperature etc. to obtain the filtered data set. Identify and correct the outliers in the filtered data set. Outliers are detected using the Z-score method, that is, calculate the mean and standard deviation of each performance parameter, and then determine whether a certain data point exceeds the set threshold (usually taking ), and the outlier detection formula is as follows:

[0050]

[0051] If , then is regarded as an outlier and needs to be corrected. The correction method uses linear interpolation, that is:

[0052]

[0053] Among them, is the corrected data point. After this process, a corrected data set is obtained. The corrected data set is normalized. Min-max normalization is adopted to map all data to the interval [0, 1], and its calculation formula is:

[0054]

[0055] Among them, and are the minimum and maximum values of this performance parameter respectively. After normalization, all performance parameters are in the same numerical range, avoiding the excessive influence of parameters with a large numerical range on subsequent calculations. All the normalized data are reconstructed into a matrix with the sampling time points as row vectors and the performance parameters as column vectors to obtain the battery performance measurement data matrix:

[0056]

[0057] Based on this data matrix, a zero-mean censored normal distribution model is constructed, and the optimal value of the noise variance is calculated using the principle of maximum entropy. The matrix is zero-meaned, that is, the mean value of each column is calculated and subtracted, so that the mean value of each column of the data matrix is zero:

[0058]

[0059] A censored region is set for the censored data points, and the abnormal parts of the data are removed through statistical analysis methods, and it is assumed that the noise follows a normal distribution . In order to determine the optimal noise variance , using the principle of maximum entropy, an entropy function is defined:

[0060]

[0061] Among them, is the probability density function of the censored normal distribution:

[0062]

[0063] Maximize to obtain the optimal value, making the entropy of the noise the largest and obtaining the noise distribution parameters with the most information balance. Based on this optimized noise parameter, a battery performance noise feature matrix is constructed, and this matrix effectively describes the noise characteristics in the battery data.

[0064] In a specific embodiment, the process of constructing a zero-mean censored normal distribution model based on the battery performance measurement data matrix, calculating the optimal value of the noise variance through the principle of maximum entropy, and obtaining the battery performance noise feature matrix may specifically include the following steps:

[0065] Perform data block division on the battery performance measurement data matrix to obtain multiple data blocks, and perform centering processing on each data block to obtain zero-mean data blocks;

[0066] Divide the censored interval for the zero-mean data blocks to obtain the censored interval, and construct a zero-mean censored normal distribution model based on the censored interval;

[0067] Use the principle of maximum entropy to construct an optimization objective function for noise variance and perform numerical optimization and solution on the zero-mean censored normal distribution model to obtain the optimal value of the noise variance;

[0068] Construct a white noise channel model based on the zero-mean censored normal distribution model and the optimal value of the noise variance, and perform noise analysis on the battery performance measurement data matrix to obtain the noise distribution characteristics;

[0069] Calculate the noise entropy values of each performance parameter for the noise distribution characteristics, construct the noise entropy values into a noise entropy feature set, and perform correlation mapping and matrix transformation operations on the noise entropy feature set and the corresponding performance parameters to obtain the battery performance noise feature matrix.

[0070] Specifically, perform data block division on the battery performance measurement data matrix. Let the battery performance measurement data matrix be , where represents the th performance parameter collected at time , such as voltage , current , temperature , internal resistance , and state of charge . This matrix is expressed as:

[0071]

[0072] where is the number of time sampling points, and the number of columns is different battery performance parameters. Divide the matrix along the time axis into blocks, and each data block contains consecutive time points, that is:

[0073]

[0074] where represents the th data block, and the size of each block is 。Center each data block, that is, calculate the mean for each parameter and subtract the mean to make the mean of the data block zero. Let the mean of a certain performance parameter in the

[0075]

[0076] th data block be:

[0077]

[0078] Then the zero-mean data block after centering is represented as: Divide the zero-mean data block into censored intervals. The censored intervals are used to identify outliers and high-noise regions in the data, which are set based on the standard deviation threshold, that is, if the absolute value of a certain data point exceeds

[0079]

[0080] times its standard deviation, then this point is considered to belong to the censored interval. Let the standard deviation be:

[0081]

[0082] Among them, usually take 3 to ensure that the censored data points are outliers. On this basis, construct a zero-mean censored normal distribution model. Assume that the censored data follows a censored normal distribution, and its probability density function is:

[0083]

[0084] Based on this model, use the principle of maximum entropy to optimize the estimated value of the noise variance . The definition of entropy is:

[0085]

[0086] To obtain the maximum entropy , use numerical optimization methods, such as gradient descent or Lagrange multiplier method, to solve:

[0087]

[0088] Get the optimal noise variance , that is, the noise variance under the maximum entropy condition. Based on the optimal noise variance, construct a white noise channel model, and represent the white noise channel as:

[0089]

[0090] Among them, is the observed data, is the real signal, is zero-mean white noise. Conduct noise analysis on the battery performance measurement data matrix, calculate the noise distribution characteristics, and define the variance of the noise as:

[0091]

[0092] Calculate the noise entropy values of each performance parameter based on the noise distribution characteristics. The noise entropy value measures the randomness of the noise and is calculated using Shannon entropy:

[0093]

[0094] where, is the probability density function of the observed data. Construct the noise entropy feature set from all the noise entropy values:

[0095]

[0096] Perform an association mapping between the noise entropy feature set and the corresponding performance parameters to construct a noise feature matrix:

[0097]

[0098] In a specific embodiment, the process of executing step S2 may specifically include the following steps:

[0099] Calculate the marginal entropy of the performance parameters in the battery performance noise feature matrix, and construct the joint entropy for the marginal entropy to obtain the performance parameter entropy value set;

[0100] Based on the performance parameter entropy value set, construct the mutual information calculation formula and calculate the mutual information for all pairs of performance parameters to obtain the mutual information matrix;

[0101] Subtract the corresponding noise entropy value from the mutual information matrix, construct the performance parameter net information feature, calculate the correlation measure and redundancy based on the performance parameter net information feature, and construct the feature evaluation function;

[0102] Execute the sequential forward selection algorithm on the feature evaluation function, select the feature that maximizes the F-value increment in each iteration, and mark the features with F-values greater than the first target value as candidate features to obtain the feature candidate set;

[0103] Conduct a correlation analysis on the feature candidate set, select the feature combinations with the absolute value of the correlation coefficient greater than the second target value to obtain the battery performance key feature set;

[0104] Construct an integrated feature metric function for the operating condition parameters and the performance degradation rate based on the battery performance key feature set, and obtain the battery degradation feature vector by density clustering of the local extreme points of the integrated feature metric function.

[0105] Specifically, based on the battery performance noise feature matrix Perform probability distribution estimation. Assume that contains samples, and each sample corresponds to performance parameters, namely:

[0106]

[0107] where represents the -th performance parameter of the -th sample. For each performance parameter , based on its probability distribution , calculate the marginal entropy. The formula is as follows:

[0108]

[0109] where represents the marginal entropy of the -th performance parameter, and represents the probability that the value is . Calculate the joint entropy of multiple performance parameters. For two performance parameters and , the joint entropy is defined as:

[0110]

[0111] where represents the joint probability distribution that the value of is and the value of . By calculating the marginal entropy and joint entropy of all performance parameters, construct the performance parameter entropy value set . Based on the performance parameter entropy value set, construct the mutual information calculation formula to measure the information correlation degree between different performance parameters. The mutual information calculation formula is as follows:

[0112]

[0113] where the mutual information reflects the correlation between the performance parameters and . The larger the value, the higher the information overlap degree between the two. Construct the mutual information matrix :

[0114]

[0115] To eliminate the influence of noise, subtract the corresponding noise entropy value from the mutual information matrix, thereby constructing the net information feature of the performance parameter. The noise entropy is calculated as follows:

[0116]

[0117] Among them, represents the probability distribution of the noise signal The performance parameter net information feature is calculated as follows:

[0118]

[0119] Calculate the correlation measure and redundancy based on the performance parameter net information feature to construct a feature evaluation function. The correlation measure The calculation formula is as follows:

[0120]

[0121] And the redundancy The calculation formula is as follows:

[0122]

[0123] Construct the feature evaluation function :

[0124]

[0125] Based on the feature evaluation function, execute the sequential forward selection algorithm. In each iteration, select the feature that maximizes the value increment, and mark the features greater than the first target value as candidate features to form a feature candidate set. Conduct a correlation analysis on the feature candidate set to select a feature combination whose absolute value of the correlation coefficient is greater than the second target value to obtain the key feature set of battery performance. The correlation coefficient is calculated as follows:

[0126]

[0127] Among them, is and 's covariance, and are respectively and 's standard deviations. Screen the feature pairs that satisfy to construct the key feature set of battery performance. Based on the key feature set of battery performance, construct an integrated feature metric function for operating condition parameters and performance decay rate to comprehensively measure the influence of different features on battery decay. The integrated feature metric function is calculated as follows:

[0128]

[0129] Among them, is the weight of the th feature. In order to identify the key decay features, the local extreme points of this function are analyzed. By the density clustering method, set the time span to , define the threshold of the number of points within the cluster , and execute the density clustering algorithm. The clustering results are as follows:

[0130]

[0131] Among them, each represents a feature cluster. By extracting the clustering center points, a battery decay feature vector is formed:

[0132]

[0133] In a specific embodiment, the process of executing the steps of constructing an integrated feature metric function of operating condition parameters and performance decay rate according to the key feature set of battery performance, and obtaining the battery decay feature vector by density clustering processing of the local extreme points of the integrated feature metric function may specifically include the following steps:

[0134] Extract the operating condition parameters from the key feature set of battery performance, calculate the performance decay rate, and set a balance factor to construct an integrated feature metric function;

[0135] Set a sliding time window for the integrated feature metric function, and calculate the local mean and standard deviation within the time window to obtain local statistical features;

[0136] Determine the local threshold condition according to the local statistical features, mark the time points that meet the local threshold condition, and obtain a set of candidate feature points;

[0137] Execute the density clustering algorithm on the set of candidate feature points with a preset time span as the clustering radius, and perform cluster division based on the point density and distance threshold to obtain the clustering result of feature points;

[0138] Based on the clustering result of feature points, select the center point of each cluster, record the time stamp, operating condition parameters, and performance parameters corresponding to the center point, obtain a sequence of key feature points, and reconstruct the sequence of key feature points into a vector form according to the time order to obtain the battery decay feature vector.

[0139] Specifically, select the parameters related to the operating condition from the key feature set of battery performance, such as charge-discharge rate, battery temperature, SOC change range, etc., and at the same time calculate the performance decay rate of the battery, and this decay rate is expressed as the capacity retention rate and the internal resistance growth rate The change situation. Assume that at time the nominal capacity of the battery is , then its relative capacity is defined as:

[0140]

[0141] where is the actually measured battery capacity at time . Similarly, the relative change rate of the battery internal resistance is expressed as:

[0142]

[0143] where is the actual internal resistance at time , is the initial internal resistance. On this basis, the balance factor is set to construct an integrated feature metric function to balance the relationship between the operating condition parameters and the performance decay rate. The integrated feature metric function is defined as:

[0144]

[0145] where is the weight coefficient of the corresponding feature, used to control the contribution of different features to the comprehensive metric, and is adjusted through an optimization method. A sliding time window is set for the integrated feature metric function to analyze the local statistical characteristics of the battery decay features. The window size is set to , and the local mean and the standard deviation are calculated within each window to obtain the statistical features within this window:

[0146]

[0147]

[0148] where reflects the feature mean within the current window, represents the feature fluctuation situation within this time window. According to the local statistical features, local threshold conditions are determined to screen out potential key time points. The local threshold is set to:

[0149]

[0150] where is the threshold factor, taking 2 or 3 to ensure that the selected time points reflect the significant features of the battery decay. If or ​​, then mark this time point as a candidate feature point to obtain a set of candidate feature points. Perform a density clustering algorithm on the set of candidate feature points to identify the decay patterns under different working conditions. Density clustering is based on the density distribution of data points, divides high-density regions into clusters, and filters out isolated points. Set a time span as the clustering radius and define the minimum number of points in a cluster to perform the density clustering algorithm. If the number of candidate feature points around a certain time point exceeds , then this point is regarded as a core point, and the neighboring points around it belong to the same cluster. This process is calculated using the following distance metric:

[0151]

[0152] If , then and belong to the same cluster to obtain the feature point clustering result :

[0153]

[0154] After obtaining the feature point clustering result, extract the center point of each cluster and record the time stamps, working condition parameters, and performance parameters corresponding to these center points. The center point is selected as the point with the highest density within the cluster or the point that best represents the overall trend in the cluster, that is, calculate the mean center within each cluster:

[0155]

[0156] Record the working condition parameters and performance parameters corresponding to this time point to form a key feature point sequence:

[0157]

[0158] Reconstruct the key feature point sequence in chronological order to form a battery decay feature vector. Let the time series of the key feature points be , then the battery decay feature vector is expressed as:

[0159]

[0160] This vector contains the time evolution information of battery decay and integrates the key performance parameters under different working conditions.

[0161] In a specific embodiment, the process of executing step S3 may specifically include the following steps:

[0162] Calculate the time interval, operating condition difference, and performance difference for adjacent feature points in the battery degradation feature vector, and perform weighted adaptive distance calculation to obtain the feature point distance matrix;

[0163] Set a connection threshold for the feature point distance matrix, mark the feature point pairs with a distance less than the connection threshold as connection point pairs, and obtain the initial connection relationship set;

[0164] Construct an adjacency matrix A based on the initial connection relationship set. The matrix element Aij takes 1 when the feature points i and j are connected, and 0 otherwise, to obtain a binary adjacency matrix;

[0165] Calculate the degree centrality and betweenness centrality for each node in the binary adjacency matrix, and perform normalization and weighted summation of the calculated degree centrality and betweenness centrality to obtain the node importance index;

[0166] Execute the Louvain community discovery algorithm on the binary adjacency matrix based on the node importance index, and perform community division through the modularity Q value optimization to obtain the network community structure;

[0167] Combine the topological features of the network community structure with the binary adjacency matrix to construct the battery degradation feature correlation network.

[0168] Specifically, define the battery degradation feature vector. Assume that the battery degradation feature vector consists of feature points, and each feature point contains time , operating condition parameters and performance parameters , then the feature vector is expressed as:

[0169]

[0170] To measure the difference between adjacent feature points and , calculate the time interval , operating condition difference and performance difference respectively. The calculation method of the time interval is as follows:

[0171]

[0172] The operating condition difference is calculated using the Euclidean distance:

[0173]

[0174] The performance difference is calculated in a similar way, also using the Euclidean distance:

[0175]

[0176] To comprehensively measure the overall distance between feature points, weighted adaptive distance calculation is performed. Define the weight parameters corresponding to the time interval, operating condition difference, and performance difference respectively. Then the weighted adaptive distance is calculated as follows:

[0177]

[0178] Among them, the value of is optimized based on experimental data, so that different features have reasonable influence weights when calculating the distance. Calculate the weighted adaptive distance between all pairs of feature points to form the feature point distance matrix :

[0179]

[0180] After obtaining the feature point distance matrix, set the connection threshold to screen out the feature point pairs with strong correlation. If , then and are regarded as connection point pairs to form the initial connection relationship set :

[0181]

[0182] Based on the initial connection relationship set, construct the adjacency matrix , where takes the value of 1 when and are connected, otherwise takes the value of 0 to form a binary adjacency matrix:

[0183]

[0184] Among them:

[0185]

[0186] After constructing the binary adjacency matrix, calculate the degree centrality and betweenness centrality of each node. The degree centrality is defined as the number of neighbors directly connected to the node:

[0187]

[0188] The betweenness centrality measures the bridging role of the node in the shortest path and is defined as:

[0189]

[0190] Among them, is the number of shortest paths from node to node , and is the number of these paths passing through node . Normalize and , and calculate the node importance index :

[0191]

[0192] where and are the normalized weights, adjusting the contributions of degree centrality and betweenness centrality in calculating importance. Based on the node importance index, perform the Louvain community discovery algorithm on the binary adjacency matrix, and optimize the community division through the modularity value. The modularity is defined as follows:

[0193]

[0194] where and are the degrees of node and node respectively, and represent the communities to which the nodes belong, takes 1 when , otherwise takes 0. By maximizing the value, obtain the optimal community division result and get the network community structure. Combine the topological features of the network community structure with the binary adjacency matrix to construct a battery degradation feature correlation network, which effectively describes the correlation relationship between different degradation modes.

[0195] In a specific embodiment, the process of executing step S4 may specifically include the following steps:

[0196] Extract the global topological features of the battery degradation feature correlation network, and construct a network global structure feature vector. The global topological features include the average shortest path length, network diameter, network density, and global clustering coefficient;

[0197] Calculate the node-level topological features of the battery degradation feature correlation network, and splice the node-level topological features with the network global structure feature vector to obtain a node feature matrix. The node-level topological features include node degree distribution, betweenness centrality, and eigenvector centrality;

[0198] Extract community-level topological features from the battery degradation feature correlation network, and combine them with the node feature matrix to obtain a multi-scale network feature vector. The community-level topological features include the number of communities, the community size distribution, and the connection strength between communities;

[0199] Input the multi-scale network feature vector and the battery degradation feature vector into the input layer of the radial basis kernel function support vector machine classifier to obtain standardized input features, and perform non-linear mapping on the standardized input features through the radial basis kernel function to obtain kernel function mapping features;

[0200] Execute the classification decision function based on the kernel function mapping features to classify the performance degradation mode of the energy storage battery, and obtain the classification result of the performance degradation mode of the energy storage battery.

[0201] Specifically, extract the global topological features of the battery degradation feature correlation network. Let the battery degradation feature correlation network be , where is the node set, representing battery degradation feature points, is the edge set, representing the correlation relationship between different degradation feature points. To describe the overall structure of this network, calculate the global topological features, including the average shortest path length, network diameter, network density, and global clustering coefficient. The average shortest path length is calculated as follows:

[0202]

[0203] where, is the number of nodes in the network, represents the shortest path length between nodes and . This metric is used to measure the overall connectivity of the network. The network diameter is calculated as follows:

[0204]

[0205] It represents the shortest path length between the two farthest nodes in the network, reflecting the expansion range of the network. The network density is calculated as follows:

[0206]

[0207] where, is the number of edges in the network. This metric is used to measure the tightness of the connection between nodes in the network. The global clustering coefficient is calculated as follows:

[0208]

[0209] where, is the node The local clustering coefficient, which represents the connectivity among the neighbors of the node, is calculated as follows:

[0210]

[0211] where is the actual number of connections among the neighbors of node , and is the degree of node (i.e., the number of directly connected neighbors). After calculating the above global topological features, they are combined into the network global structure feature vector :

[0212]

[0213] Calculate the node-level topological features of the battery degradation feature correlation network and splice them with the global structure feature vector to form the node feature matrix. The node-level topological features include node degree distribution, betweenness centrality, and eigenvector centrality. The node degree is calculated as follows:

[0214]

[0215] where is an element of the adjacency matrix. If there is a connection between nodes and , then , otherwise . The betweenness centrality is calculated as follows:

[0216]

[0217] where is the number of shortest paths from node to node , and is the number of these paths passing through node . The eigenvector centrality is calculated as follows:

[0218]

[0219] where is the eigenvector, is the adjacency matrix, and is the largest eigenvalue. This centrality is used to measure the influence of a node in the entire network. Combine the above node-level topological features to form the node feature matrix :

[0220]

[0221] Extract community-level topological features and combine them with the node feature matrix to form a multi-scale network feature vector. The community-level topological features include the number of communities, the community size distribution, and the connection strength between communities. Let be the th community, and the number of communities is defined as:

[0222]

[0223] The community size distribution is calculated as follows:

[0224]

[0225] The connection strength between communities is calculated as follows:

[0226]

[0227] Combine the community-level topological features to obtain the multi-scale network feature vector :

[0228]

[0229] Take the multi-scale network feature vector and the battery degradation feature vector as inputs and input them into the input layer of the radial basis kernel function support vector machine. Perform standardization processing to ensure that different features have the same scale. The standardization formula is as follows:

[0230]

[0231] where is the mean, and is the standard deviation. Use the radial basis kernel function to perform non-linear mapping:

[0232]

[0233] where is a hyperparameter used to control the smoothness of the high-dimensional mapping. After obtaining the kernel function mapping features, execute the classification decision function:

[0234]

[0235] where are the coefficients of the support vectors, is the class label, and is the bias term. Based on the classification decision function, classify the performance degradation mode of the energy storage battery and obtain the final classification result.

[0236] The battery performance degradation test method of the energy storage battery in the embodiment of the present invention has been described above. Next, the battery performance degradation test device of the energy storage battery in the embodiment of the present invention will be described. Please refer to Figure 2 , an embodiment of the battery performance degradation test device of the energy storage battery in the embodiment of the present invention includes:

[0237] A construction module, configured to construct a battery performance measurement data matrix during the charge and discharge process of the energy storage battery, and calculate the optimal value of the noise variance through a zero-mean censored normal distribution model to obtain a battery performance noise feature matrix;

[0238] A calculation module, configured to calculate the mutual information of performance parameters based on the battery performance noise feature matrix to obtain a battery performance key feature set, and construct a battery decay feature vector of the operating condition parameters and the performance decay rate;

[0239] A creation module, configured to extract the weighted adaptive distances of the time interval, the operating condition difference, and the performance difference according to the battery decay feature vector, and create a battery decay feature association network;

[0240] A classification module, configured to classify the performance decay mode based on the battery decay feature association network and the battery decay feature vector to obtain a classification result of the performance decay mode of the energy storage battery.

[0241] Through the collaborative cooperation of the above-mentioned various components, the present invention models the measurement noise by adopting a zero-mean censored normal distribution model, and combines the maximum entropy principle to calculate the optimal value of the noise variance, realizing the effective elimination of the noise in the measurement data. At the same time, Butterworth low-pass filtering and outlier processing are adopted to ensure the data quality; through the feature dimension reduction method based on the mutual information theory and the sequential forward selection algorithm, and the integrated feature metric function of the operating condition parameters and the performance decay rate, the battery performance decay features are comprehensively extracted; the weighted adaptive distance calculation and the community structure adjacency matrix construction method are introduced, combined with the Louvain community discovery algorithm, to reveal the internal connection between the performance decay features; the multi-scale network feature extraction method is adopted to depict the performance decay features from three levels of global, node, and community, and the radial basis kernel function support vector machine classifier is used to realize the accurate classification of the performance decay mode.

[0242] Refer to Figure 3 , an embodiment of the present invention also provides a computer device, which may be a server, and its internal structure may be as Figure 3As shown in the figure. The computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected via a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the above method is implemented.

[0243] Those skilled in the art can understand that Figure 3 the structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0244] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described in detail here.

[0245] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs, etc., which can store program codes.

[0246] The above is the case. The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of various embodiments of the present invention.

Claims

1. A method for testing battery performance attenuation of an energy storage battery, characterized in that: The method comprises: A battery performance measurement data matrix of an energy storage battery during the charging and discharging process is constructed, and a zero-mean censored normal distribution model is constructed based on the battery performance measurement data matrix, and the optimal value of the noise variance is calculated by the maximum entropy principle to obtain a battery performance noise characteristic matrix; the battery performance noise characteristic matrix is ​​obtained specifically including: data block division of the battery performance measurement data matrix to obtain multiple data blocks, and centering each data block to obtain a zero-mean data block; censoring interval division of the zero-mean data block to obtain a censored interval, and constructing a zero-mean censored normal distribution model according to the censored interval ; Use the maximum entropy principle to construct the noise variance optimization objective function and numerically optimize the zero-mean censored normal distribution model to obtain the optimal value of the noise variance; construct a white noise channel model based on the zero-mean censored normal distribution model and the optimal value of the noise variance, and perform noise analysis on the battery performance measurement data matrix to obtain noise distribution characteristics; calculate the noise entropy value of each performance parameter based on the noise distribution characteristics, construct the noise entropy value into a noise entropy feature set, and perform association mapping and matrix transformation operations on the noise entropy feature set and the corresponding performance parameters to obtain a battery performance noise feature matrix; Based on the battery performance noise characteristic matrix, the mutual information of performance parameters is calculated to obtain a key characteristic set of battery performance, and a battery attenuation characteristic vector of operating condition parameters and performance attenuation rate is constructed; Extracting the weighted adaptive distance of the time interval, the operating condition difference and the performance difference according to the battery attenuation feature vector, and creating a battery attenuation feature association network; Based on the battery attenuation feature association network and the battery attenuation feature vector, performance attenuation mode classification is performed to obtain a classification result of the energy storage battery performance attenuation mode.

2. The battery performance attenuation testing method of the energy storage battery according to claim 1, characterized in that: The method of constructing a battery performance measurement data matrix of an energy storage battery during the charging and discharging process includes: Collect data on the voltage, current, temperature, internal resistance and state of charge of the energy storage battery during the charging and discharging process to obtain the original data sequence; Performing segment processing on the original data sequence to obtain a time series data set, and performing signal processing on the time series data set to obtain a filtered data set; Identifying and correcting outliers in the filtered data set to obtain a corrected data set, and normalizing all performance parameters in the corrected data set to obtain a normalized data set; The normalized data set is reconstructed into a matrix using the sampling time points as row vectors and the performance parameters as column vectors to obtain a battery performance measurement data matrix.

3. The battery performance attenuation testing method of the energy storage battery according to claim 1, characterized in that: The method of calculating the mutual information of performance parameters based on the battery performance noise characteristic matrix to obtain a key characteristic set of battery performance and constructing a battery attenuation characteristic vector of operating condition parameters and performance attenuation rate includes: Calculating the marginal entropy of the performance parameters in the battery performance noise characteristic matrix, and performing joint entropy construction on the marginal entropy to obtain a performance parameter entropy value set; Constructing a mutual information calculation formula based on the performance parameter entropy value set, and performing mutual information calculation on all performance parameter pairs to obtain a mutual information matrix; Subtracting the corresponding noise entropy value from the mutual information matrix, and constructing a performance parameter net information feature, calculating a correlation measure and a redundancy according to the performance parameter net information feature, and constructing a feature evaluation function; Execute a sequential forward selection algorithm on the feature evaluation function, select the feature with the largest F value increment in each iteration, and mark the features with F values ​​greater than the first target value as candidate features to obtain a feature candidate set; Performing correlation analysis on the feature candidate set, selecting a feature combination whose absolute value of the correlation coefficient is greater than a second target value, and obtaining a battery performance key feature set; An integrated feature metric function of operating parameters and performance attenuation rate is constructed according to the key feature set of battery performance, and local extreme points of the integrated feature metric function are processed by density clustering to obtain a battery attenuation feature vector.

4. The battery performance attenuation testing method of the energy storage battery according to claim 3 is characterized in that: The method of constructing an integrated characteristic metric function of operating condition parameters and performance attenuation rate according to the key characteristic set of battery performance, and processing the local extreme points of the integrated characteristic metric function by density clustering to obtain a battery attenuation characteristic vector includes: Extracting operating parameters from the battery performance key feature set and calculating the performance decay rate, and setting a balance factor to construct an integrated feature measurement function; Setting a sliding time window for the integrated feature metric function, and calculating the local mean and standard deviation within the time window to obtain local statistical features; Determine a local threshold condition according to the local statistical feature, mark the time points that meet the local threshold condition, and obtain a set of candidate feature points; Execute a density clustering algorithm on the candidate feature point set with a preset time span as the clustering radius, perform cluster division based on point density and distance threshold, and obtain a feature point clustering result; Based on the characteristic point clustering result, the center point of each cluster is selected, and the timestamp, operating parameters and performance parameters corresponding to the center point are recorded to obtain a key characteristic point sequence, and the key characteristic point sequence is reconstructed into a vector form in chronological order to obtain a battery attenuation characteristic vector.

5. The battery performance attenuation testing method of the energy storage battery according to claim 1, characterized in that: The step of extracting the weighted adaptive distance of the time interval, the operating condition difference and the performance difference according to the battery attenuation feature vector and creating a battery attenuation feature association network includes: Calculating the time interval, operating condition difference and performance difference of adjacent characteristic points in the battery attenuation characteristic vector, and performing weighted adaptive distance calculation to obtain a characteristic point distance matrix; Setting a connection threshold for the feature point distance matrix, marking feature point pairs whose distance is less than the connection threshold as connection point pairs, and obtaining an initial connection relationship set; Based on the initial connection relationship set, an adjacency matrix A is constructed. The matrix element A ij When feature points i and j are connected, take 1, otherwise take 0, and get a binary adjacency matrix; Calculating the degree centrality and betweenness centrality of each node in the binary adjacency matrix, and normalizing and weighted summing the calculated degree centrality and the betweenness centrality to obtain a node importance index; Based on the node importance index, the Louvain community discovery algorithm is executed on the binary adjacency matrix, and the community is divided by optimizing the modularity Q value to obtain the network community structure; The topological features of the network community structure are combined with the binary adjacency matrix to construct a battery attenuation feature association network.

6. The battery performance attenuation testing method of the energy storage battery according to claim 1, characterized in that: The performance attenuation mode classification is performed based on the battery attenuation feature association network and the battery attenuation feature vector to obtain a classification result of the energy storage battery performance attenuation mode, including: Extracting the global topological features of the battery attenuation feature association network and constructing a global structural feature vector of the network, wherein the global topological features include an average shortest path length, a network diameter, a network density, and a global clustering coefficient; Calculating node-level topological features for the battery attenuation feature association network, and concatenating the node-level topological features with the network global structure feature vector to obtain a node feature matrix, wherein the node-level topological features include node degree distribution, betweenness centrality, and feature vector centrality; Extracting community-level topological features from the battery attenuation feature association network, and combining them with the node feature matrix to obtain a multi-scale network feature vector, wherein the community-level topological features include the number of communities, community size distribution, and connection strength between communities; Inputting the multi-scale network feature vector and the battery attenuation feature vector into the input layer of the radial basis kernel function support vector machine classifier to obtain standardized input features, and performing nonlinear mapping on the standardized input features through the radial basis kernel function to obtain kernel function mapping features; A classification decision function is executed based on the kernel function mapping feature to classify the performance attenuation mode of the energy storage battery to obtain a classification result of the performance attenuation mode of the energy storage battery.

7. A battery performance attenuation test device for an energy storage battery, characterized in that: A battery performance degradation test method for an energy storage battery according to any one of claims 1 to 6, wherein the battery performance degradation test device for the energy storage battery comprises: A construction module is used to construct a battery performance measurement data matrix of the energy storage battery during the charging and discharging process, and calculate the optimal value of the noise variance through a zero-mean censored normal distribution model to obtain a battery performance noise characteristic matrix; A calculation module, used to calculate the mutual information of performance parameters based on the battery performance noise characteristic matrix, obtain a key characteristic set of battery performance, and construct a battery attenuation characteristic vector of operating condition parameters and performance attenuation rate; Creating a module for extracting a weighted adaptive distance of a time interval, a working condition difference, and a performance difference according to the battery attenuation feature vector, and creating a battery attenuation feature association network; A classification module is used to classify the performance attenuation mode based on the battery attenuation feature association network and the battery attenuation feature vector to obtain a classification result of the energy storage battery performance attenuation mode.

8. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the battery performance attenuation test method of the energy storage battery described in any one of claims 1 to 6 when executing the computer program.

Images