Power battery capacity abnormity identification method and system, electronic equipment and storage medium

Abstract

The invention provides a power battery capacity abnormity identification method and system, electronic equipment and a storage medium, and relates to the technical field of battery detection, and the method comprises the steps: obtaining cloud battery data of a power battery, carrying out the data preprocessing, obtaining a battery standard data set, constructing a battery capacity increment curve of each single battery, and obtaining a battery capacity increment curve of each single battery; extracting curve peak characteristics in the battery capacity increment curve; obtaining off-line battery data of the power battery to determine key curve characteristics; constructing a principal component regression model according to the key curve features, and performing regression fitting estimation by adopting the principal component regression model to obtain key feature weights; obtaining the battery monomer capacity of each battery monomer according to the curve peak characteristic and the key characteristic weight; and calculating a battery monomer deviation degree according to the battery monomer capacity, and determining a battery monomer with abnormal capacity based on the battery monomer deviation degree. According to the invention, robustness of power battery capacity abnormity identification is improved, and precision and real-time performance of battery monomer capacity prediction are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of battery detection, and in particular to a method, system, electronic device and storage medium for identifying abnormal capacity of a power battery. Background Art

[0002] As the new energy vehicle industry booms, power batteries, as core components, play a crucial role in all types of new energy vehicles, including pure electric vehicles, hybrid vehicles, and fuel cell vehicles. In pure electric vehicles, which represent the largest share of the new energy vehicle market, power battery costs account for approximately 30%-50% of the total vehicle cost, and their service life is generally shorter than the overall service life of the new energy vehicle. This characteristic has made power battery performance evaluation and health management a key focus of the industry.

[0003] The core factor affecting the lifespan of power batteries lies in their capacity decay characteristics. Over long-term use, the available capacity of power batteries gradually decreases due to multiple factors, including electrochemical reactions, material aging, and physical loss. It is worth noting that power battery systems typically consist of dozens to hundreds of battery cells connected in series and parallel. Due to complex factors such as differences in battery manufacturing processes, changes in operating environments, and charge and discharge strategies, the overall performance of a battery system is often limited by the performance of its worst-performing cell, resulting in a significant "barrel effect." Currently, degradation detection of power batteries in traditional new energy vehicles primarily relies on offline testing, which can only be performed at specific time intervals and cannot achieve real-time monitoring of battery status, making it difficult to detect potential problems in a timely manner. Therefore, how to accurately and accurately assess the capacity decay and evolution trends of battery cells in a power battery system in real time is an urgent issue that needs to be addressed. Summary of the Invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method, system, electronic device and storage medium for identifying power battery capacity anomalies, which effectively solve the problem of being unable to perform real-time and accurate assessment of the capacity attenuation and evolution trend of battery cells in a power battery system.

[0005] In a first aspect, the present invention provides a method for identifying abnormal capacity of a power battery, the method comprising:

[0006] Obtaining cloud battery data of the power battery and performing data preprocessing on the cloud battery data to obtain a battery standard data set;

[0007] constructing a battery capacity increment curve for each battery cell according to the battery standard data set, and extracting a curve peak feature in the battery capacity increment curve;

[0008] Obtaining offline battery data of the power battery, and determining key curve characteristics based on the offline battery data;

[0009] Constructing a principal component regression model based on the key curve characteristics, and performing regression fitting estimation using the principal component regression model to obtain key feature weights;

[0010] Obtaining a battery cell capacity of each of the battery cells according to the curve peak feature and the key feature weight;

[0011] A battery cell deviation is calculated according to the battery cell capacity, and a battery cell with abnormal capacity is determined based on the battery cell deviation.

[0012] In an optional embodiment, the method further comprises:

[0013] The battery cluster capacity is calculated according to the battery cell capacity, and a maintenance strategy is determined based on the battery cluster capacity and the battery cell capacity.

[0014] In an optional embodiment, constructing a battery capacity increment curve for each battery cell according to the battery standard data set and extracting a curve peak feature in the battery capacity increment curve includes:

[0015] Extracting charging data of each battery cell under target operating conditions according to the battery standard data set;

[0016] Performing Gaussian regression fitting on the charging data to obtain fitted charging data;

[0017] performing differential calculation on the fitted charging data to obtain differential data;

[0018] constructing the battery capacity increment curve according to the differential data;

[0019] A target peak is selected according to the battery capacity increment curve, and a curve peak feature of the target peak is extracted.

[0020] In an optional embodiment, the acquiring offline battery data of the power battery and determining the key curve characteristics according to the offline battery data includes:

[0021] Performing a laboratory simulation test on the power battery to obtain the offline battery data;

[0022] Constructing an offline capacity increment curve for each battery cell of the power battery according to the offline battery data;

[0023] Peak features are extracted according to the offline capacity increment curve, and correlation analysis is performed on the peak features to obtain the key curve features.

[0024] In an optional embodiment, constructing a principal component regression model according to the key curve characteristics, and using the principal component regression model to perform regression fitting estimation to obtain key feature weights includes:

[0025] Dividing the key curve features into a training set and a test set, and performing standardization on the training set and the test set to obtain a standard training set and a standard test set;

[0026] Constructing a principal component regression model based on the standard training set and verifying it using the standard test set;

[0027] The principal component regression model is used to calculate the principal components to obtain the key feature weights.

[0028] In an optional embodiment, calculating the battery cell deviation according to the battery cell capacity, and determining the battery cell with abnormal capacity based on the battery cell deviation includes:

[0029] Calculate the average capacity and capacity standard deviation based on the capacity of the battery cells in the battery cluster;

[0030] Calculate the battery cell deviation according to the average capacity, the capacity standard deviation and the battery cell capacity;

[0031] If the battery cell deviation of the target battery cell is less than a preset deviation threshold, the target battery cell is determined to be the battery cell with abnormal capacity.

[0032] In an optional embodiment, the calculating the battery cluster capacity according to the battery cell capacity, and determining the maintenance strategy based on the battery cluster capacity and the battery cell capacity, includes:

[0033] Obtaining a first battery cell capacity of a first battery cell and a second battery cell capacity of a second battery cell in a battery cluster, where the first battery cell is a battery cell corresponding to a highest battery cell voltage in the battery cluster at the end of charging, and the second battery cell is a battery cell corresponding to a lowest battery cell voltage in the battery cluster at the start of charging;

[0034] Calculating a battery capacity difference based on the first battery cell capacity and the second battery cell capacity;

[0035] Calculating the battery cluster capacity according to the battery capacity difference and the first battery cell capacity;

[0036] If the battery cluster capacity is equal to the minimum battery cell capacity in the battery cluster, replacing the target battery cell corresponding to the minimum battery cell capacity;

[0037] If the capacity of the battery cluster is less than the minimum battery cell capacity, a single cell balancing operation is performed on the battery cluster.

[0038] In a second aspect, the present invention provides a power battery capacity abnormality identification system, the system comprising:

[0039] A data processing module is used to obtain cloud battery data of the power battery and perform data preprocessing on the cloud battery data to obtain a battery standard data set;

[0040] a feature extraction module, configured to construct a battery capacity increment curve for each battery cell based on the battery standard data set, and extract a curve peak feature from the battery capacity increment curve;

[0041] a feature determination module, configured to obtain offline battery data of the power battery and determine key curve features based on the offline battery data;

[0042] A model building module is used to build a principal component regression model based on the key curve characteristics, and use the principal component regression model to perform regression fitting estimation to obtain key feature weights;

[0043] a capacity estimation module, configured to obtain a battery cell capacity of each of the battery cells according to the peak feature of the curve and the weight of the key feature;

[0044] The abnormality identification module is used to calculate the battery cell deviation according to the battery cell capacity, and determine the battery cell with abnormal capacity based on the battery cell deviation.

[0045] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for identifying abnormal power battery capacity as described in the first aspect of the present invention.

[0046] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for identifying abnormal power battery capacity as described in the first aspect of the present invention.

[0047] The power battery capacity anomaly identification method, system, electronic device, and storage medium provided by the present invention collect battery data in real time through the cloud, calculate the mapping relationship between incremental capacity and voltage, generate a high-resolution battery capacity increment curve, extract the curve peak characteristic parameters from the battery capacity increment curve, and use principal component analysis to reduce the dimensionality of high-dimensional features, retaining the main information while reducing computational complexity and effectively filtering out noise interference, thereby improving the robustness of power battery capacity anomaly identification. It can dynamically adapt to different battery types, aging stages, and operating conditions, achieving accurate and real-time capacity prediction. At the same time, combined with the changes in the health status of the power battery system, it clarifies the main influencing factors of power battery capacity anomalies and formulates corresponding maintenance strategies, which is of great significance for maintaining the health of the power battery system and improving its service life. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0049] Figure 1 This is a first schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention;

[0050] Figure 2 This is a second schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention;

[0051] Figure 3 1 is a schematic diagram of a battery capacity increment curve of a battery cell in an embodiment of the present invention;

[0052] Figure 4 This is a third schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention;

[0053] Figure 5 This is a fourth schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention;

[0054] Figure 6 Schematic diagram of principal component regression analysis in an embodiment of the present invention;

[0055] Figure 7 This is a fifth schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention;

[0056] Figure 8 This is a sixth schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention;

[0057] Figure 9 This is a schematic diagram of the structure of a power battery capacity abnormality identification system provided by an embodiment of the present invention;

[0058] Figure 10 It is a structural diagram of an electronic device provided by an embodiment of the present invention.

[0059] Description of main component symbols:

[0060] 200. Power battery capacity anomaly identification system; 210. Data processing module; 220. Feature extraction module; 230. Feature determination module; 240. Model building module; 250. Capacity estimation module; 260. Anomaly identification module; 300. Electronic device; 310. Processor; 320. Communication interface; 330. Memory; 340. Communication bus. DETAILED DESCRIPTION

[0061] To make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be further clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. It should be noted that the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0062] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature identified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0063] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. The terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit the present invention. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0064] The core factor affecting the lifespan of power batteries lies in their capacity decay characteristics. Over long-term use, the available capacity of power batteries gradually decreases due to multiple factors, including electrochemical reactions, material aging, and physical loss. It is worth noting that power battery systems typically consist of dozens to hundreds of battery cells connected in series and parallel. Due to complex factors such as differences in battery manufacturing processes, changes in operating environments, and charge and discharge strategies, the overall performance of a battery system is often limited by the performance of its worst-performing cell, resulting in a significant "barrel effect." Currently, degradation detection of power batteries in traditional new energy vehicles primarily relies on offline testing, which can only be performed at specific time intervals and cannot achieve real-time monitoring of battery status, making it difficult to detect potential problems in a timely manner. Therefore, how to accurately and accurately assess the capacity decay and evolution trends of battery cells in a power battery system in real time is an urgent issue that needs to be addressed.

[0065] Example 1

[0066] The embodiment of the present invention provides a method for identifying abnormal capacity of a power battery, which effectively solves the problem of being unable to perform real-time and accurate assessment of the capacity attenuation and evolution trend of battery cells in a power battery system. Figure 1 This is a first schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention. Figure 1 As shown, the method includes the following steps:

[0067] S100: Obtain cloud battery data of the power battery, and perform data preprocessing on the cloud battery data to obtain a battery standard data set.

[0068] In an embodiment of the present invention, the vehicle's battery management system uploads collected battery data to the cloud via an onboard terminal and stores it in the cloud, obtaining cloud-based battery data. This cloud-based battery data includes, but is not limited to, total battery voltage, total current, maximum cell voltage, minimum cell voltage, voltage list data, temperature list data, charge capacity, discharge capacity, and accumulated mileage.

[0069] Data preprocessing primarily involves data cleaning, standardization, and event segmentation. Data cleaning involves filling missing values using methods like the median and mode, deleting duplicate data, and identifying and removing outliers using the interquartile range (IQR) statistic. Standardization unifies the data format, for example, by processing data according to time series. Event segmentation is based on the operating status of the power battery or vehicle, defining charging, discharging, and resting states.

[0070] S200 : constructing a battery capacity increment curve for each battery cell according to a battery standard data set, and extracting a peak feature of the battery capacity increment curve.

[0071] Figure 2 This is a second schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention. Figure 2 As shown, extracting the peak feature of the curve specifically includes the following steps:

[0072] S210 : Extracting charging data of each battery cell under target operating conditions according to a battery standard data set.

[0073] In this embodiment of the present invention, to avoid the impact of fluctuations in the battery capacity increment curve on the battery capacity estimation results, it is necessary to select an appropriate target operating condition and extract the charging data for each battery cell from the battery standard data set based on the target operating condition. Optionally, a charging condition can be selected in which the charging current is stable and the temperature is relatively stable, with the starting SOC of charging less than 30% and the ending SOC of charging = 100%. Stable charging current means that the charging rate is between 0.2C and 0.6C, and the current fluctuation does not exceed ±10% of the mean. Relatively stable temperature means that the charging environment temperature is between 15-35 degrees Celsius.

[0074] S220 , performing Gaussian regression fitting on the charging data to obtain fitted charging data.

[0075] Due to the randomness and complexity of on-board power battery charging, it is difficult to directly extract effective features. Therefore, we choose to reconstruct the charging data based on Gaussian process regression to extract efficient capacity estimation features. Specifically, some charging data are reconstructed using a Gaussian process regression model. The similarity between data points is defined by a kernel function. The kernel function is then maximized to optimize the hyperparameters of the kernel function. The Gaussian model is trained based on the kernel function and the charging data. The trained Gaussian model can then be used for regression prediction. For each missing time point, the Gaussian model predicts the corresponding voltage or current value, extracts the mean from the predicted distribution, or selects a value within the confidence interval as the completion value based on application requirements. The completion value is used to supplement the missing data within a pre-selected fixed voltage range to obtain fitted charging data.

[0076] S230 , performing differential calculation on the fitted charging data to obtain differential data.

[0077] Based on the Gaussian regression fitted charging data, the data is compressed by a fixed voltage interval in a fixed voltage range. For example, the fixed voltage interval of the iron-lithium power battery is 1mV, and the fixed voltage interval of the ternary power battery is 3mV. The fixed voltage interval is used as the microelement for differential calculation, and the calculation formula is as follows:

[0078]

[0079] In the above formula, i≥2, dQ / dV represents the i ,V i ) differential data, Q iRepresents the capacity in the i-th fixed voltage interval, V i represents the voltage in the i-th fixed voltage interval, Q i-1 Represents the capacity in the i-1th fixed voltage interval, V i-1 Represents the voltage in the i-1th fixed voltage interval.

[0080] S240: Construct a battery capacity increment curve based on the differential data.

[0081] In the embodiment of the present invention, a battery capacity increment curve is constructed with voltage as the horizontal axis and differential data dQ / dV as the vertical axis. Figure 3 FIG. 1 is a schematic diagram of a battery capacity increment curve of a battery cell in an embodiment of the present invention. Figure 3 As shown in the figure, by analyzing the battery capacity increment curve, the electrochemical characteristics of the battery, such as the phase change peak position and peak area, can be extracted, thereby evaluating the battery's capacity attenuation, internal resistance change and aging mechanism.

[0082] S250 , selecting a target peak according to the battery capacity increment curve, and extracting a curve peak feature of the target peak.

[0083] There are multiple peaks in the battery capacity increment curve, and each peak represents a specific voltage platform. The peak characteristics are directly related to the phase change of the electrode material and are highly sensitive to battery aging. However, because the starting SOC of the vehicle-side battery charging is uncertain, and as the power battery system ages, the inconsistency between battery clusters increases, some information of Peak 1 is easily lost. Therefore, Peaks 2 and 3 are selected as target peaks to extract the curve peak characteristics of the target peaks. The curve peak characteristics include but are not limited to peak value, peak area, peak charging time, peak voltage, peak height, peak left and right slope, peak left and right height, peak left and right voltage, peak height mean, peak height standard deviation, peak kurtosis, and peak skewness.

[0084] In an embodiment of the present invention, data segmentation is performed on the peak of the battery capacity increment curve to obtain 5 data points, namely position 1, position 2, position 3, position 4 and position 5, where position 1 is peak 1, and the corresponding voltage is shifted to the left by 0.02V; position 2 is peak 1, and the corresponding voltage is shifted to the right by 0.02V; position 3 is peak 2, and the corresponding voltage is shifted to the left by 0.015V; position 4 is peak 2, and the corresponding voltage is shifted to the right by 0.015V; position 5 is peak 2, and the corresponding voltage is shifted to the right by 0.02V.

[0085] Then, the peak characteristics of the curve are obtained based on the data of each data point, where the peak height represents the incremental capacity value of the highest point in different intervals. Taking Peak 2 as an example, the highest point value of dQ / dV in positions 3 and 4 is selected as the peak height of Peak 2. The peak value represents the voltage corresponding to the peak height, and the peak area represents the charging capacity between different positions. For example, the area of Peak 2 is the charging capacity corresponding to position 4 minus the charging capacity corresponding to position 3. The peak charging time represents the charging time difference between different positions, such as the charging time of Peak 2 is the charging time corresponding to position 4 minus the charging time corresponding to position 3. The left slope of the peak represents the ratio of the incremental capacity change value at the peak height and the adjacent point on the left to the voltage change value. For example, the calculation formula of the left slope of Peak 2 is as follows:

[0086]

[0087] In the above formula, k1 represents the left slope of peak 2, IC 峰2 represents the incremental capacity value at peak 2, IC3 represents the incremental capacity value at position 3, V 峰2 represents the voltage at peak 2, and V3 represents the voltage at position 3.

[0088] The peak right slope represents the ratio of the incremental capacity change to the voltage change between the peak height and the adjacent point on the right. For example, the calculation formula for the peak 2 right slope is as follows:

[0089]

[0090] In the above formula, k2 represents the right slope of peak 2, IC 峰2 represents the incremental capacity value at peak 2, IC4 represents the incremental capacity value at position 4, V 峰2 represents the voltage at peak 2, and V4 represents the voltage at position 4.

[0091] The mean peak height represents the average of the incremental capacity values across different peaks. For example, the mean peak height of peak 2 represents the average of the incremental capacity values from positions 3 to 4. The standard deviation of the peak height represents the standard deviation of the incremental capacity values across different peaks. For example, the mean peak height of peak 2 represents the standard deviation of the incremental capacity values from positions 3 to 4. Skewness refers to the degree of peak offset and is a measure of the asymmetry of the data distribution. Peak-to-peak kurtosis refers to the maximum difference between two adjacent peaks and is a measure of the kurtosis of the data distribution.

[0092] S300: Obtain offline battery data of the power battery, and determine key curve characteristics based on the offline battery data.

[0093] Figure 4 FIG3 is a third schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention. Figure 4 As shown in FIG, determining the key curve characteristics specifically includes the following steps:

[0094] S310: Perform a laboratory simulation test on the power battery to obtain offline battery data.

[0095] In an embodiment of the present invention, various operating conditions under actual use conditions are simulated in a laboratory to obtain offline battery data of the power battery under different environments (e.g., temperature, humidity), loads (e.g., charge and discharge current), and usage time. This offline battery data includes, but is not limited to, battery cell test capacity, total battery test voltage, total test current, maximum cell test voltage, minimum cell test voltage, voltage list test data, temperature list test data, test charge capacity, and test discharge capacity.

[0096] S320: Construct an offline capacity increment curve for each battery cell of the power battery according to the offline battery data.

[0097] In the embodiment of the present invention, an offline capacity increment curve of each battery cell of the power battery is constructed based on the offline battery data in the same manner as steps S210 to S240 .

[0098] S330 , extracting peak features according to the offline capacity increment curve, performing correlation analysis on the peak features, and obtaining key curve features.

[0099] The peak characteristics of each battery cell can be extracted based on the offline capacity increment curve. The peak characteristics include but are not limited to peak value, peak area, peak charging time, peak voltage, peak height, left and right slopes of peak, left and right heights of peak, left and right voltages of peak height, peak height mean, peak height standard deviation, peak kurtosis and peak skewness.

[0100] In the embodiment of the present invention, since the offline battery data obtained based on laboratory testing includes the battery cell test capacity of each battery cell, the Pearson correlation coefficient between the peak characteristics of each battery cell and the battery cell test capacity can be calculated to perform correlation analysis. The calculation formula of the Pearson correlation coefficient is as follows:

[0101]

[0102] In the above formula, r represents the Pearson correlation coefficient, n represents the total number of eigenvalues, and X i represents the i-th eigenvalue of the peak feature, The mean of the eigenvalues representing the peak feature, Y i represents the i-th eigenvalue of the battery cell test capacity, Indicates the mean of the characteristic values of the test capacity of the battery cell.

[0103] The peak features of all battery cells whose Pearson correlation coefficients are greater than a preset coefficient threshold are determined as key curve features.

[0104] S400: Construct a principal component regression model based on key curve features, perform regression fitting estimation using the principal component regression model, and obtain key feature weights.

[0105] In order to solve the problem of multicollinearity between key curve features, the principal component regression method is selected to reconstruct the features and generate a few unrelated variables, namely principal components, so that the principal components can reflect as much information of the original variables as possible, and the key feature weights are determined. The battery cell capacity is estimated through regression analysis. Figure 5 FIG4 is a fourth schematic diagram of a method for identifying abnormal power battery capacity according to an embodiment of the present invention. Figure 5 As shown in the figure, capacity estimation specifically includes the following steps:

[0106] S410 , dividing the key curve features into a training set and a test set, and performing standardization processing on the training set and the test set to obtain a standard training set and a standard test set.

[0107] In the embodiment of the present invention, the peak features of the curve are divided into a training set and a test set, and then the training set and the test set are standardized to obtain a standard training set and a standard test set. The formula for the standardization process is as follows:

[0108]

[0109] In the above formula, Z represents the standard eigenvalue, X represents the eigenvalue in the training set or test set, P represents the mean of the eigenvalue in the training set or test set, and S represents the standard deviation of the eigenvalue in the training set or test set.

[0110] S420. Construct a principal component regression model based on the standard training set and verify it using the standard test set.

[0111] In the embodiment of the present invention, the covariance matrix is calculated based on the standard eigenvalues in the standard training set, and the elements σ of the covariance matrix are ij represents the standard eigenvalue Z i and Z j The covariance of is calculated as follows:

[0112]

[0113] In the above formula, σ ij represents the standard eigenvalue Z i and Z j covariance, n represents the number of samples, Z ki represents the value of the i-th variable of the k-th sample, Z kj represents the j-th variable value of the k-th sample, u i and u j Represent the standard eigenvalue Z i and Z jThe mean of .

[0114] Then perform eigendecomposition on the covariance matrix to obtain multiple eigenvalues λ1, ..., λ n , and the corresponding eigenvectors v1,…,v n , where the eigenvalue represents the variance of the corresponding principal component, and the eigenvector represents the direction of the corresponding principal component. Then, based on the size of the eigenvalue, we select the first k largest eigenvalues and their corresponding eigenvectors. Assuming that we select the first three principal components, the corresponding eigenvectors are v1, v2, and v3. The three selected eigenvectors are used to construct the dimensionality reduction matrix W = [v1v2v3].

[0115] Project the data in the standard training set into the new feature space to obtain the reduced dimensionality data. The dimensionality reduction formula is as follows:

[0116] Xpca=Z×W

[0117] In the above formula, X pca Represents the feature data after dimensionality reduction, Z represents the standard eigenvalue, and W represents the dimensionality reduction matrix.

[0118] Then the principal component interpretation is performed. The variance explanation ratio represents the contribution rate of each principal component to the total variance. The calculation formula of the variance explanation ratio is as follows:

[0119]

[0120] In the above formula, K represents the variance explanation ratio, λ i represents the eigenvalue corresponding to the i-th principal component, and p represents the total number of principal components. It represents the sum of all principal component eigenvalues, that is, the total variance.

[0121] Finally, the principal component score is calculated. The principal component score is the projection value of each sample in the principal component direction, that is, the feature data X after dimensionality reduction. pca As a new feature, it is used in regression analysis to obtain the initial principal component regression model.

[0122] Optionally, the initial principal component regression model is verified using key curve features in a standard test set to obtain a principal component regression model.

[0123] S430: Calculate the principal components using a principal component regression model to obtain key feature weights.

[0124] In an embodiment of the present invention, the principal component features that can cover more than 95% of the information of the original features are selected to output the principal component kernel load matrix. For example, five curve peak features, namely peak area, peak charging data, peak height, peak left slope and peak right slope, are selected to construct five principal components, namely PC1, PC2, PC3, PC4 and PC5. The principal component kernel load matrix between the principal components and the original features is shown in Table 1.

[0125] Table 1. Schematic diagram of principal component core loading matrix

[0126] Peak area Peak charging time Peak height Peak left slope Peak right slope PC1 0.45 0.45 0.45 0.44 -0.44 PC2 -0.48 -0.48 0.08 0.70 -0.21 PC3 -0.13 -0.13 -0.18 -0.42 -0.87 PC4 0.22 0.22 -0.87 0.38 -0.07 PC5 -0.71 0.71 0.00 -0.00 0.00

[0127] The principal component feature weight is used to measure the relative importance of the original variable in a specific principal component and can be calculated using the absolute value of the standardized load or the square normalization of the load.

[0128] Figure 6 Schematic diagram of principal component regression analysis in an embodiment of the present invention, such as Figure 6 As shown in the figure, the cumulative variance explanation ratio changes with the increase in the number of principal components. The first three principal components, namely PC1, PC2 and PC3, can explain 99% of the information of the five curve peak features, namely peak area, peak charging data, peak height, peak left slope and peak right slope. Therefore, PC1, PC2 and PC3 are used to replace the original five curve peak features to achieve dimensionality reduction and eliminate multicollinearity between variables.

[0129] S500 : Obtain the battery cell capacity of each battery cell according to the curve peak feature and the key feature weight.

[0130] In the embodiment of the present invention, based on the trained principal component regression model and key feature weights, the peak feature of the curve of each battery cell is selected as the principal component feature for regression fitting to estimate the battery cell capacity of each battery cell. The battery cell capacity is numbered according to the coding sequence of the battery cell. For example, the battery cell capacity corresponding to battery cell No. 1 is C1, and the battery cell capacity corresponding to battery cell No. n is C. n , and so on.

[0131] S600 , calculating a battery cell deviation according to the battery cell capacity, and determining a battery cell with abnormal capacity based on the battery cell deviation.

[0132] Figure 7 This is a fifth schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention. Figure 7 As shown, the identification of battery cells with abnormal capacity specifically includes the following steps:

[0133] S610 : Calculate the average capacity and the capacity standard deviation according to the capacities of the battery cells in the battery cluster.

[0134] A battery cluster is an integrated unit composed of multiple battery cells connected in series and parallel. The average capacity and capacity standard deviation can be calculated based on the capacity of all the battery cells in the battery cluster. The formula for calculating the average capacity is as follows:

[0135]

[0136] In the above formula, C avg represents the average capacity of the battery cluster, n represents the total number of battery cells in the battery cluster, C i Represents the battery cell capacity of the i-th battery cell in the battery cluster.

[0137] The calculation formula for capacity standard deviation is as follows:

[0138]

[0139] In the above formula, C std Indicates the capacity standard deviation of the battery cluster.

[0140] S620 , obtaining a battery cell deviation degree by calculation according to the average capacity, the capacity standard deviation, and the battery cell capacity.

[0141] In the embodiment of the present invention, the calculation formula of the battery cell deviation is as follows:

[0142]

[0143] In the above formula, devi i The cell deviation of the i-th cell is represented by a positive or negative sign, which indicates whether the cell capacity is higher or lower than the average capacity.

[0144] S630: If the battery cell deviation of the target battery cell is less than the preset deviation threshold, determine that the target battery cell is a battery cell with abnormal capacity.

[0145] In the embodiment of the present invention, if there is a devi i If the value of the capacity of the battery cell is less than the preset deviation threshold, the battery cell of the i-th battery cell is determined to be a battery cell with abnormal capacity, and the corresponding battery cluster is a battery cluster with abnormal capacity. Optionally, the preset deviation threshold can be set to -3.

[0146] As an optional implementation of the embodiment of the present invention, the method further includes the following steps:

[0147] S700 : Calculate the battery cluster capacity according to the battery cell capacity, and determine a maintenance strategy based on the battery cluster capacity and the battery cell capacity.

[0148] Figure 8This is a sixth schematic diagram of the process of the method for identifying abnormal power battery capacity provided by an embodiment of the present invention. Figure 8 As shown in Figure 2, the determination of the maintenance strategy specifically includes the following steps:

[0149] S710 : Obtain a first battery cell capacity of a first battery cell and a second battery cell capacity of a second battery cell in a battery cluster.

[0150] In the embodiment of the present invention, the first battery cell is the battery cell corresponding to the highest cell voltage in the battery cluster at the end of charging, and the capacity of the first battery cell is C max The second battery cell is the battery cell with the lowest cell voltage in the battery cluster at the start of charging. The capacity of the second battery cell is C min .

[0151] S720 , calculating a battery capacity difference based on the first battery cell capacity and the second battery cell capacity.

[0152] Typically, Peak 1 in the battery capacity increment curve represents the beginning of lithium insertion at the negative electrode. At this time, the peak is less affected by impedance. The battery capacity difference can be determined by the time difference between different battery cells reaching the peak value of Peak 1. Assuming that the time taken for the first battery cell to reach the peak value of Peak 1 is x, and the time taken for the second battery cell to reach the peak value of Peak 1 is y, the battery capacity difference is calculated as follows:

[0153]

[0154] In the above formula, SOC 1-2 represents the battery capacity difference, I represents the average current, x represents the time taken by the first battery cell to reach the peak value of peak 1, y represents the time taken by the second battery cell to reach the peak value of peak 1, C max Indicates the capacity of the first battery cell, C min Indicates the capacity of the second battery cell.

[0155] S730: Calculate the battery cluster capacity based on the battery capacity difference and the first battery cell capacity. In this embodiment of the present invention, the battery cluster capacity is calculated using the following formula:

[0156] C pack =C max ×(1+SOC 1-2 )

[0157] In the above formula, C pack Indicates the battery cluster capacity.

[0158] S740: If the battery cluster capacity is equal to the minimum battery cell capacity in the battery cluster, replace the target battery cell corresponding to the minimum battery cell capacity.

[0159] In an embodiment of the present invention, the battery cluster capacity is equal to the capacity of the smallest battery cell in the battery cluster, indicating that the battery cluster capacity depends on the battery cell with the shortest capacity. Replacing the battery cell with the smallest capacity or the module in which it is located can improve the health of the battery cluster.

[0160] S750: If the battery cluster capacity is less than the minimum battery cell capacity, perform a single cell balancing operation on the battery cluster.

[0161] In an embodiment of the present invention, if the battery cluster capacity is less than the capacity of the smallest battery cell in the cluster, it indicates that inconsistency between battery cells is affecting the full capacity of the battery cluster, and the battery cluster needs to be balanced. Specifically, the battery cell corresponding to the lowest charging starting voltage is identified and a low current is applied to this battery cell until its voltage matches the next lowest voltage. This improves the consistency between battery cells and enhances the health of the battery cluster.

[0162] The method for identifying abnormal power battery capacity provided by an embodiment of the present invention collects battery data in real time through the cloud, calculates the mapping relationship between incremental capacity and voltage, generates a high-resolution incremental battery capacity curve, extracts peak characteristic parameters from the curve, and uses principal component analysis to reduce the dimensionality of high-dimensional features. This method retains key information while reducing computational complexity and effectively filtering out noise interference, thereby improving the robustness of abnormal power battery capacity identification. Furthermore, by combining the changes in the health status of the power battery system, the method identifies the main influencing factors of abnormal power battery capacity and formulates corresponding maintenance strategies, which is of great significance for maintaining the health of the power battery system and improving its service life.

[0163] Example 2

[0164] Based on the same technical concept as the above method embodiment 1, the embodiment of the present invention provides a power battery capacity abnormality identification system. Figure 9 FIG. 1 is a schematic diagram of a power battery capacity abnormality identification system according to an embodiment of the present invention. Figure 9 As shown, the power battery capacity abnormality identification system 200 includes:

[0165] The data processing module 210 is used to obtain cloud battery data of the power battery and perform data preprocessing on the cloud battery data to obtain a battery standard data set;

[0166] A feature extraction module 220 is configured to construct a battery capacity increment curve for each battery cell based on a battery standard data set, and extract a peak feature from the battery capacity increment curve;

[0167] A feature determination module 230 is configured to obtain offline battery data of the power battery and determine key curve features based on the offline battery data;

[0168] A model building module 240 is used to build a principal component regression model based on key curve features, and use the principal component regression model to perform regression fitting estimation to obtain key feature weights;

[0169] a capacity estimation module 250 for obtaining the battery cell capacity of each battery cell based on the curve peak feature and the key feature weight;

[0170] The abnormality identification module 260 is configured to calculate the battery cell deviation according to the battery cell capacity, and determine the battery cell with abnormal capacity based on the battery cell deviation.

[0171] The power battery capacity anomaly identification system provided by the embodiment of the present invention can dynamically adapt to different battery types, aging stages and usage conditions, and achieve accurate and real-time capacity prediction.

[0172] It can be understood that the implementation of the power battery capacity abnormality identification method described in the above embodiment 1 is also applicable to this embodiment and can achieve the same technical effect, so it will not be repeated here.

[0173] Example 3

[0174] Based on the same concept, an embodiment of the present invention further provides an electronic device, Figure 10 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention, such as Figure 10 As shown, the electronic device 300 may include: a processor 310, a communication interface 320, a memory 330 and a communication bus 340, wherein the processor 310, the communication interface 320, and the memory 330 communicate with each other via the communication bus 340. The processor 310 may call the logic instructions in the memory 330 to execute the steps of the power battery capacity abnormality identification method described in the above embodiments. For example, it includes:

[0175] S100, obtaining cloud battery data of the power battery, and performing data preprocessing on the cloud battery data to obtain a battery standard data set;

[0176] S200, constructing a battery capacity increment curve for each battery cell based on a battery standard data set, and extracting a peak feature in the battery capacity increment curve;

[0177] S300, obtaining offline battery data of the power battery, and determining key curve characteristics based on the offline battery data;

[0178] S400, constructing a principal component regression model based on key curve characteristics, performing regression fitting estimation using the principal component regression model, and obtaining key feature weights;

[0179] S500, obtaining the battery cell capacity of each battery cell according to the curve peak feature and the key feature weight;

[0180] S600 , calculating a battery cell deviation according to the battery cell capacity, and determining a battery cell with abnormal capacity based on the battery cell deviation.

[0181] The processor 310 may be a central processing unit (CPU). 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, or a combination of the above chips.

[0182] In addition, the logic instructions in the above-mentioned memory 330 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0183] The memory 330 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely located relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0184] Example 4

[0185] Based on the same concept, an embodiment of the present invention further provides a computer-readable storage medium storing a computer program, the computer program including at least one code segment that can be executed by a main control device to control the main control device to implement the steps of the power battery capacity abnormality identification method described in the above embodiments. For example, the steps include:

[0186] S100, obtaining cloud battery data of the power battery, and performing data preprocessing on the cloud battery data to obtain a battery standard data set;

[0187] S200, constructing a battery capacity increment curve for each battery cell based on a battery standard data set, and extracting a peak feature in the battery capacity increment curve;

[0188] S300, obtaining offline battery data of the power battery, and determining key curve characteristics based on the offline battery data;

[0189] S400, constructing a principal component regression model based on key curve characteristics, performing regression fitting estimation using the principal component regression model, and obtaining key feature weights;

[0190] S500, obtaining the battery cell capacity of each battery cell according to the curve peak feature and the key feature weight;

[0191] S600 , calculating a battery cell deviation according to the battery cell capacity, and determining a battery cell with abnormal capacity based on the battery cell deviation.

[0192] Based on the same technical concept, an embodiment of the present invention further provides a computer program, which, when executed by a main control device, is used to implement the above method embodiment.

[0193] In summary, the power battery capacity anomaly identification method, system, electronic device, and storage medium provided by the present invention collect battery data in real time through the cloud, calculate the mapping relationship between incremental capacity and voltage, generate a high-resolution battery capacity increment curve, extract the curve peak characteristic parameters from the battery capacity increment curve, and use principal component analysis to reduce the dimensionality of high-dimensional features, retaining the main information while reducing computational complexity and effectively filtering out noise interference, thereby improving the robustness of power battery capacity anomaly identification, and being able to dynamically adapt to different battery types, aging stages, and operating conditions, achieving accurate and real-time capacity prediction. At the same time, combined with the changes in the health status of the power battery system, the main influencing factors of power battery capacity anomalies are clarified and corresponding maintenance strategies are formulated, which is of great significance for maintaining the health of the power battery system and improving its service life.

[0194] References to "embodiments" herein mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0195] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0196] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for identifying abnormal power battery capacity, characterized in that: The method comprises: Obtaining cloud battery data of the power battery and performing data preprocessing on the cloud battery data to obtain a battery standard data set; constructing a battery capacity increment curve for each battery cell according to the battery standard data set, and extracting a curve peak feature in the battery capacity increment curve; Obtaining offline battery data of the power battery, and determining key curve characteristics based on the offline battery data; Constructing a principal component regression model based on the key curve characteristics, and performing regression fitting estimation using the principal component regression model to obtain key feature weights; Obtaining a battery cell capacity of each of the battery cells according to the curve peak feature and the key feature weight; A battery cell deviation is calculated according to the battery cell capacity, and a battery cell with abnormal capacity is determined based on the battery cell deviation.

2. The method for identifying abnormal power battery capacity according to claim 1, characterized in that: The method further comprises: The battery cluster capacity is calculated according to the battery cell capacity, and a maintenance strategy is determined based on the battery cluster capacity and the battery cell capacity.

3. The method for identifying abnormal power battery capacity according to claim 1, characterized in that: The step of constructing a battery capacity increment curve for each battery cell according to the battery standard data set and extracting a curve peak feature in the battery capacity increment curve includes: Extracting charging data of each battery cell under target operating conditions according to the battery standard data set; Performing Gaussian regression fitting on the charging data to obtain fitted charging data; performing differential calculation on the fitted charging data to obtain differential data; constructing the battery capacity increment curve according to the differential data; A target peak is selected according to the battery capacity increment curve, and a curve peak feature of the target peak is extracted.

4. The method for identifying abnormal power battery capacity according to claim 1, characterized in that: The acquiring of offline battery data of the power battery and determining key curve characteristics according to the offline battery data include: Performing a laboratory simulation test on the power battery to obtain the offline battery data; Constructing an offline capacity increment curve for each battery cell of the power battery according to the offline battery data; Peak features are extracted according to the offline capacity increment curve, and correlation analysis is performed on the peak features to obtain the key curve features.

5. The method for identifying abnormal power battery capacity according to claim 4, characterized in that: The step of constructing a principal component regression model according to the key curve characteristics, performing regression fitting estimation using the principal component regression model, and obtaining key feature weights includes: Dividing the key curve features into a training set and a test set, and performing standardization on the training set and the test set to obtain a standard training set and a standard test set; Constructing a principal component regression model based on the standard training set and verifying it using the standard test set; The principal component regression model is used to calculate the principal components to obtain the key feature weights.

6. The method for identifying abnormal power battery capacity according to claim 1, characterized in that: The step of calculating the battery cell deviation according to the battery cell capacity and determining the battery cell with abnormal capacity based on the battery cell deviation includes: Calculate the average capacity and capacity standard deviation based on the capacity of the battery cells in the battery cluster; Calculate the battery cell deviation according to the average capacity, the capacity standard deviation and the battery cell capacity; If the battery cell deviation of the target battery cell is less than a preset deviation threshold, the target battery cell is determined to be the battery cell with abnormal capacity.

7. The method for identifying abnormal power battery capacity according to claim 1, characterized in that: The calculating the battery cluster capacity according to the battery cell capacity and determining the maintenance strategy based on the battery cluster capacity and the battery cell capacity includes: Obtaining a first battery cell capacity of a first battery cell and a second battery cell capacity of a second battery cell in a battery cluster, where the first battery cell is a battery cell corresponding to a highest battery cell voltage in the battery cluster at the end of charging, and the second battery cell is a battery cell corresponding to a lowest battery cell voltage in the battery cluster at the start of charging; Calculating a battery capacity difference based on the first battery cell capacity and the second battery cell capacity; Calculating the battery cluster capacity according to the battery capacity difference and the first battery cell capacity; If the battery cluster capacity is equal to the minimum battery cell capacity in the battery cluster, replacing the target battery cell corresponding to the minimum battery cell capacity; If the capacity of the battery cluster is less than the minimum battery cell capacity, a single cell balancing operation is performed on the battery cluster.

8. A power battery capacity abnormality identification system, characterized in that: The system comprises: A data processing module is used to obtain cloud battery data of the power battery and perform data preprocessing on the cloud battery data to obtain a battery standard data set; a feature extraction module, configured to construct a battery capacity increment curve for each battery cell based on the battery standard data set, and extract a curve peak feature from the battery capacity increment curve; a feature determination module, configured to obtain offline battery data of the power battery and determine key curve features based on the offline battery data; A model building module is used to build a principal component regression model based on the key curve characteristics, and use the principal component regression model to perform regression fitting estimation to obtain key feature weights; a capacity estimation module, configured to obtain a battery cell capacity of each of the battery cells according to the peak feature of the curve and the weight of the key feature; The abnormality identification module is used to calculate the battery cell deviation according to the battery cell capacity, and determine the battery cell with abnormal capacity based on the battery cell deviation.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the computer program to implement the method for identifying abnormal power battery capacity according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for identifying abnormal power battery capacity according to any one of claims 1 to 7 is implemented.

Images