Power battery management method and system based on data drive

Abstract

The present invention discloses a data-driven power battery management method and system, which specifically relates to the technical field of battery degradation monitoring, and is used to solve the problem in the prior art that it is difficult to monitor the hidden degradation of batteries under microcirculation conditions in real time; the present invention analyzes the microcirculation energy spectrum through a transient spectrum clustering algorithm to identify the risks of spectrum concentration anomalies and structural degradation; analyzes the battery internal resistance offset trend to evaluate the impact of internal resistance fluctuations on the granulation of active materials; determines whether there is a microcirculation abnormality risk based on spectrum anomalies and internal resistance fluctuations; when there is an abnormal risk, analyzes the stress accumulation condition through tensor decomposition and singular value decomposition, and evaluates the attenuation gradient of the electrode polarization signal in combination with variational mode decomposition and recursive neural network to determine the overall degradation trend of the battery; the present invention can dynamically optimize the management of power batteries, extend service life, and improve safety and reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of battery degradation monitoring, and more specifically, to a data-driven power battery management method and system. Background Art

[0002] In actual operation, the power battery of a hybrid vehicle is often in a micro-circulation state of shallow charge and discharge, rather than deep charge and discharge. This micro-circulation state is different from deep cycle and will have a unique impact on the internal materials of the battery. It may cause the active materials inside the battery to gradually granulate and induce structural degradation, thereby shortening the service life of the battery. In the existing technology, there is a lack of in-depth discussion on the material granulation and structural degradation phenomena under the micro-circulation state.

[0003] It is difficult to effectively identify and evaluate the material degradation trend caused by microcirculation in the existing technology, which will cause hidden damage to accumulate during the use of the battery, resulting in a gradual decline in battery performance and a shortened service life.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a data-driven power battery management method and system to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The data-driven power battery management method includes the following steps:

[0008] The microcirculation energy spectrum distribution is analyzed by transient spectrum clustering algorithm to evaluate the potential risk of structural degradation caused by spectrum concentration anomalies.

[0009] Analyze the deviation trend of battery internal resistance and evaluate the potential impact of internal resistance fluctuation on active material granulation;

[0010] Based on the potential risk of structural degradation caused by spectrum concentration anomalies and the potential impact of internal resistance fluctuations on active material granulation, determine whether the power battery has a risk of microcirculation anomalies;

[0011] When there is a risk of abnormal microcirculation in the power battery, the stress accumulation tensor of the battery cell is analyzed through tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration; the attenuation gradient of the electrode polarization signal is analyzed through variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure caused by granulation acceleration;

[0012] By comprehensively analyzing the degree of damage to the material structure caused by stress concentration and the degree of degradation of the electrode structure caused by accelerated granulation, the overall degradation trend of the power battery under microcirculation conditions is evaluated; and the power battery management is optimized based on this overall degradation trend.

[0013] In a preferred embodiment, the microcirculation energy spectrum distribution is analyzed by a transient spectrum clustering algorithm to evaluate the potential risk of abnormal spectrum concentration to structural degradation, specifically:

[0014] Decompose the energy spectrum of the power battery in the microcirculation state and extract the amplitude and frequency characteristics of the spectrum;

[0015] Divide the extracted amplitude features and frequency features into different frequency bands according to the frequencies, and generate characteristic data of the frequency bands;

[0016] The characteristic data of each frequency band are hierarchically clustered by multi-scale decomposition method to analyze the fluctuation pattern and change rate of each frequency band;

[0017] Calculate the mutation points and abnormal intervals in the spectrum and identify the spectrum energy deviation under the microcirculation state;

[0018] Based on the spectrum energy offset and hierarchical clustering results, the potential risk of spectrum concentration anomalies to structural degradation is evaluated.

[0019] In a preferred embodiment, the potential risk of structural degradation caused by spectrum concentration anomaly is evaluated based on the spectrum energy offset and the hierarchical clustering result, specifically:

[0020] Calculate the structural degradation risk value, the expression is: ;in, is the structural degradation risk value, represents the total number of clustering levels for hierarchical clustering, Indicates The weight factor of each clustering level, Indicates The fluctuation pattern of the cluster level, Indicates The rate of change of each cluster level;

[0021] Set the structural degradation risk threshold and compare the structural degradation risk value with the structural degradation risk threshold:

[0022] When the structural degradation risk value is greater than the structural degradation risk threshold, it is determined that the potential risk of the spectrum concentration anomaly to structural degradation is high; when the structural degradation risk value is less than or equal to the structural degradation risk threshold, it is determined that the potential risk of the spectrum concentration anomaly to structural degradation is normal.

[0023] In a preferred embodiment, the deviation trend of the battery internal resistance is analyzed to evaluate the potential impact of the internal resistance fluctuation on the granulation of the active material, specifically:

[0024] Collect the internal resistance data of the power battery in the microcirculation state and extract the offset characteristics through the internal resistance curve;

[0025] The Gaussian process regression model is used to model the internal resistance offset characteristics, and the Gaussian process regression is used to predict the future offset of the internal resistance. The formula is as follows: ;in, For time The internal resistance value at the moment, is the mean function, is the covariance function, Indicates another point in time. represents a Gaussian process;

[0026] Calculate the granulation risk value, the expression is: ;in, is the granular risk value, For the analysis period, Indicates the rate of change of internal resistance;

[0027] Set the granular risk threshold and compare the granular risk value with the granular risk threshold:

[0028] When the particle size risk value is greater than the particle size risk threshold, it is determined that the internal resistance deviation of the power battery poses a risk of material degradation; when the particle size risk value is less than or equal to the particle size risk threshold, it is determined that the internal resistance fluctuation of the power battery does not pose a risk of material degradation.

[0029] In a preferred embodiment, based on the potential risk of structural degradation caused by abnormal spectrum concentration and the potential impact of internal resistance fluctuation on active material granulation, it is determined whether the power battery has a risk of microcirculation abnormality, specifically:

[0030] When the potential risk of abnormal spectrum concentration to structural degradation is normal, and the fluctuation of internal resistance of the power battery does not pose a risk to material degradation, it is determined that there is no risk of abnormal microcirculation in the power battery; otherwise, it is determined that there is a risk of abnormal microcirculation in the power battery.

[0031] In a preferred embodiment, the stress accumulation tensor of the battery cell is analyzed by tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration, specifically:

[0032] Collect stress data of power batteries under abnormal microcirculation conditions, and apply filtering methods to denoise and smooth the data;

[0033] The preprocessed stress data are organized according to three-dimensional spatial coordinates to construct a third-order stress tensor;

[0034] Perform tensor decomposition on the stress tensor to obtain stress components in different directions, revealing the local concentration of stress distribution in each direction;

[0035] Based on the results of tensor decomposition, the singular value decomposition method is used to extract the singular values ​​in the stress tensor;

[0036] The stress damage influence index is calculated by the maximum value among the singular values ​​to quantify the damage degree of stress concentration to the material structure.

[0037] In a preferred embodiment, the attenuation gradient of the electrode polarization signal is analyzed by variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure due to granulation acceleration, specifically:

[0038] Collect electrode polarization signals when the microcirculation of the power battery is abnormal;

[0039] The electrode polarization signal is decomposed into different modes using the variational mode decomposition method to obtain the frequency components and attenuation characteristics of each mode.

[0040] A recursive neural network model is constructed based on the frequency components and attenuation characteristics of the modes to learn the granulation acceleration trends of different modes;

[0041] The historical electrode polarization signals are imported into the recurrent neural network model for training, and the recurrent neural network model is used to predict the future changes of the attenuation gradient and obtain the granulation attenuation rate;

[0042] The output of the recursive neural network model is used as the prediction result of the attenuation gradient of each mode, and the particle cumulative degradation index is calculated to quantify the degree of degradation of the electrode structure caused by particle acceleration.

[0043] In a preferred embodiment, the particle cumulative degradation index is calculated to quantify the degree of degradation of the electrode structure due to particle acceleration, specifically:

[0044] The granulation decay rate of each mode is used to calculate the particle cumulative degradation index, which is expressed as: ;in, is the particle cumulative degradation index, is the total number of modes, is the modality number, is the total time period for evaluation, It is The granular decay rate of each mode.

[0045] In a preferred embodiment, the overall degradation trend of the power battery under microcirculation is evaluated by comprehensively analyzing the degree of damage to the material structure caused by stress concentration and the degree of degradation of the electrode structure caused by granulation acceleration; and the power battery management is optimized based on the overall degradation trend, specifically:

[0046] The stress damage influence index and the particle cumulative degradation index are normalized, and weight coefficients are assigned to the normalized stress damage influence index and the particle cumulative degradation index, respectively, to calculate the overall degradation trend index;

[0047] Set the overall degradation trend threshold; compare the overall degradation trend index with the overall degradation trend threshold:

[0048] When the overall degradation trend index is greater than the overall degradation trend threshold, it is determined that the overall degradation trend of the power battery is obvious; at this time, the power battery management is optimized;

[0049] When the overall degradation trend index is less than or equal to the overall degradation trend threshold, it is determined that the overall degradation trend of the power battery is normal.

[0050] On the other hand, the present invention provides a data-driven power battery management system, including a spectrum clustering analysis module, an internal resistance shift analysis module, a cycle risk determination module, a stress accumulation analysis module, a polarization degradation analysis module, and an overall degradation assessment module;

[0051] Spectrum clustering analysis module: Analyze the microcirculation energy spectrum distribution through transient spectrum clustering algorithm to evaluate the potential risk of structural degradation caused by spectrum concentration anomaly;

[0052] Internal resistance deviation analysis module: Analyze the deviation trend of battery internal resistance and evaluate the potential impact of internal resistance fluctuation on active material granulation;

[0053] Cycle risk determination module: Based on the potential risk of structural degradation caused by spectrum concentration anomalies and the potential impact of internal resistance fluctuations on active material granulation, it determines whether the power battery has microcirculation anomaly risks;

[0054] Stress accumulation analysis module: When there is a risk of abnormal microcirculation in the power battery, the stress accumulation tensor of the battery cell is analyzed through tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration;

[0055] Polarization degradation analysis module: When there is a risk of abnormal microcirculation in the power battery, the attenuation gradient of the electrode polarization signal is analyzed through variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure caused by granulation acceleration;

[0056] Overall degradation assessment module: By comprehensively analyzing the degree of damage to the material structure caused by stress concentration and the degree of degradation of the electrode structure caused by granulation acceleration, the overall degradation trend of the power battery under microcirculation conditions is assessed; and the power battery management is optimized based on the overall degradation trend.

[0057] The technical effects and advantages of the data-driven power battery management method and system of the present invention are as follows:

[0058] 1. The present invention uses a transient spectrum clustering algorithm to analyze the microcirculation energy spectrum distribution, and can identify the potential impact of spectrum concentration anomalies on structural degradation under microcirculation conditions. By analyzing the trend of battery internal resistance deviation, the potential impact of internal resistance fluctuations on the granulation of active materials is evaluated, thereby effectively making up for the shortcomings of traditional deep cycle research on battery degradation assessment under microcirculation conditions. Combined with spectrum anomaly analysis and internal resistance deviation monitoring, the present invention can determine in real time whether there is a risk of microcirculation anomalies in the power battery, and achieve early warning of hidden battery degradation, which helps to accurately identify subtle degradation trends during use and ensure battery reliability.

[0059] 2. When it is determined that the battery has a risk of microcirculation abnormality, the present invention further analyzes the stress accumulation state of the battery cell through tensor decomposition and singular value decomposition, and evaluates the degree of damage to the material structure caused by stress concentration. At the same time, the analysis of the attenuation gradient of the electrode polarization signal by combining variational mode decomposition and recursive neural network is used to effectively evaluate the impact of accelerated granulation on the degradation of the electrode structure. Through a comprehensive analysis of stress concentration damage and granulation degradation, the present invention can accurately evaluate the overall degradation trend of the power battery, and optimize the battery management strategy based on this trend, dynamically adjust the charging and discharging parameters, thereby extending the service life of the battery and improving its operating safety and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 This is a schematic diagram of a data-driven power battery management method according to the present invention;

[0061] Figure 2 It is a schematic diagram of the structure of the data-driven power battery management system of the present invention. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments 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 creative work are within the scope of protection of the present invention.

[0063] Example 1

[0064] Figure 1 The present invention provides a data-driven power battery management method, which includes the following steps:

[0065] The microcirculation energy spectrum distribution was analyzed by transient spectrum clustering algorithm to evaluate the potential risk of structural degradation caused by spectrum concentration abnormalities.

[0066] The deviation trend of the battery internal resistance is analyzed to evaluate the potential impact of internal resistance fluctuation on the particle formation of active materials.

[0067] Based on the potential risk of structural degradation caused by spectrum concentration anomalies and the potential impact of internal resistance fluctuations on the granulation of active materials, it is determined whether the power battery has the risk of microcirculation anomalies.

[0068] When there is a risk of microcirculation abnormality in the power battery, the stress accumulation tensor of the battery cell is analyzed through tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration; the attenuation gradient of the electrode polarization signal is analyzed through variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure caused by granularization acceleration.

[0069] By comprehensively analyzing the degree of damage to the material structure caused by stress concentration and the degree of degradation of the electrode structure caused by accelerated granulation, the overall degradation trend of the power battery under microcirculation conditions is evaluated; and the power battery management is optimized based on this overall degradation trend.

[0070] The present invention uses a data-driven method to monitor and analyze the microcirculation status of the power battery in real time to accurately assess the degradation risk of the battery; first, the collected battery energy spectrum data, internal resistance change data, electrode polarization signal and other multi-source data are used to deeply analyze the data through transient spectrum clustering, tensor decomposition, recursive neural network and other algorithms. These algorithms can automatically extract potential structural damage and granular degradation characteristics in the microcirculation process, and predict the degradation trend of the battery in real time. Data-driven can continuously optimize decisions during the battery operation process, dynamically adjust charging and discharging strategies and management measures, thereby improving the battery life and safety.

[0071] The microcirculation energy spectrum distribution is analyzed by transient spectrum clustering algorithm to evaluate the potential risk of structural degradation caused by spectrum concentration anomaly, specifically:

[0072] Decompose the energy spectrum of the power battery in the microcirculation state and extract the amplitude and frequency characteristics of the spectrum:

[0073] When the power battery is in a shallow charge and discharge state, due to frequent micro-cycle charge and discharge operations, the energy spectrum of the battery will show specific distribution characteristics.

[0074] For the energy spectrum data of the power battery, the following formula is used to express its signal distribution: ;in, Indicates frequency The power density at For time The energy signal of the power battery in the microcirculation state at all times, is an imaginary unit ( ), Represents the frequency variable (used to describe the power density distribution of the energy spectrum at different frequencies).

[0075] Through this decomposition method, the two core features of the spectrum, namely the amplitude feature and the frequency feature, can be obtained.

[0076] The amplitude feature represents the power distribution of the spectrum at different frequencies and is used to identify the energy concentration in different frequency bands. The frequency feature characterizes the distribution position of the spectrum in each frequency band for subsequent frequency band division.

[0077] The extracted amplitude features and frequency features are divided into different frequency bands according to the frequency, and the characteristic data of the frequency bands are generated:

[0078] The extracted amplitude features and frequency features are divided according to frequency, the spectrum is divided into several frequency bands, and characteristic data of each frequency band is generated to facilitate further hierarchical clustering analysis.

[0079] Assume that the decomposed frequency features are divided into ,in Representative The center frequency of the frequency band, the amplitude characteristic represents the energy amplitude of the frequency band at the corresponding frequency. For each frequency band, the characteristic data is expressed as: ;in, For the The characteristic data of the frequency band, It is The amplitude characteristics of the frequency band, is the number of the frequency band.

[0080] The characteristic data of each frequency band constitute the input data of hierarchical clustering to further analyze the fluctuation pattern and change rate of each frequency band under the microcirculatory state.

[0081] The characteristic data of each frequency band are hierarchically clustered by multi-scale decomposition method to analyze the fluctuation pattern and change rate of each frequency band:

[0082] In the hierarchical clustering analysis stage, the characteristic data of each frequency band is hierarchically clustered using a multi-scale decomposition method to identify the correlation between frequency band characteristics. The clustering process considers the fluctuation pattern and change rate between frequency bands, and further explores the high-risk spectrum distribution patterns that may cause battery structure degradation.

[0083] Obtain a data set of characteristic data of the frequency band, and use a multi-scale decomposition method to decompose the frequency band characteristics into different clustering levels to construct a hierarchical clustering result. Suppose the clustering result is: ;in, Indicates The clustering level in the clustering level contains frequency band characteristics; In hierarchical clustering, The first clustering level frequency band characteristics; is the number of the cluster level in hierarchical clustering; Represents the number of frequency band features contained in a clustering level.

[0084] Each clustering level groups similar frequency band features to identify the distribution pattern of spectrum energy among frequency bands, and then analyze the fluctuation pattern and change rate of each frequency band.

[0085] The output of hierarchical clustering includes the fluctuation pattern of the frequency bands in each clustering level and their changing rate.

[0086] Calculate the mutation points and abnormal intervals in the spectrum and identify the spectrum energy deviation under the microcirculation state:

[0087] The identification of mutation points and abnormal intervals helps to determine whether the spectrum energy fluctuates significantly in a specific frequency band, providing characteristic information for subsequent hierarchical clustering.

[0088] Assume that the energy distribution of the spectrum changes with time as expressed by , use the following formula to calculate the change rate of the mutation point: ;in, It's time The rate of change of the mutation point at the moment.

[0089] If the change rate of a mutation point exceeds its corresponding preset threshold, it is marked as a mutation point, its position is recorded and included in the abnormal interval.

[0090] The abnormal interval is defined as the time period when the energy change rate exceeds the normal range. The specific formula is: ;in, It is the preset threshold corresponding to the change rate of the mutation point.

[0091] After identifying the mutation points and abnormal intervals, they are used as a supplement to the frequency band feature data to assist in identifying risk features in hierarchical clustering.

[0092] Among them, the setting of the preset threshold corresponding to the change rate of the mutation point is based on the spectrum change law in the historical data of the power battery. First, the spectrum change rate under a large number of normal working conditions is statistically analyzed to determine its average value and standard deviation; then the threshold is set to three times the standard deviation of the normal change rate in order to distinguish significantly abnormal changes.

[0093] According to the spectrum energy offset and hierarchical clustering results, the potential risk of spectrum concentration anomaly to structural degradation is evaluated:

[0094] Calculate the structural degradation risk value, the expression is: ;in, is the structural degradation risk value, represents the total number of clustering levels for hierarchical clustering, Indicates The weight factor of each clustering level, Indicates The fluctuation pattern of the cluster level, Indicates The rate of change of the clustering level.

[0095] in, Used to indicate the degree of impact of the group on the overall risk, the weight factor can be calculated based on the concentration of the band in the high-frequency range, or set by expert experience. Used to represent the variation characteristics of the spectrum in this hierarchical clustering level. Obtained through hierarchical clustering analysis. It is used to indicate the speed of change of the spectrum in the clustering level. It is calculated by analyzing the change of the frequency band characteristics in the clustering level over time.

[0096] It can be seen from the expression that the larger the structural degradation risk value, the higher the potential risk of spectrum concentration anomaly on structural degradation. The structural degradation risk value calculates the comprehensive impact of the fluctuation pattern and change rate in each clustering level by weighted calculation. A larger structural degradation risk value reflects that the degree of abnormal concentration of spectrum energy in the high-frequency range is more significant, and the fluctuation and change rate of the frequency band characteristics are higher. This state usually indicates that the possibility of battery structural degradation is stronger.

[0097] Set the structural degradation risk threshold and compare the structural degradation risk value with the structural degradation risk threshold:

[0098] When the structural degradation risk value is greater than the structural degradation risk threshold, it is judged that the potential risk of the spectrum concentration anomaly to structural degradation is high, indicating that the spectrum concentration anomaly may pose a high risk to structural degradation, the abnormal concentration of spectrum energy in the high-frequency range is more significant, and the fluctuation and change rate of the frequency band characteristics are relatively high.

[0099] When the structural degradation risk value is less than or equal to the structural degradation risk threshold, the potential risk of spectrum concentration anomaly on structural degradation is judged to be normal, which means that the impact of spectrum concentration anomaly on structural degradation is still within the normal range, and the concentration of spectrum energy and the fluctuation of frequency band characteristics are at a safe level.

[0100] Among them, the structural degradation risk threshold is a preset reference value used to determine whether the structural degradation risk is within a safe range. The threshold is usually set based on historical data and experience values, reflecting the maximum risk level allowed for the battery structure under normal working conditions.

[0101] The deviation trend of the battery internal resistance is analyzed to evaluate the potential impact of internal resistance fluctuations on the granulation of active materials, specifically:

[0102] Collect the internal resistance data of the power battery in the microcirculation state, extract the internal resistance offset characteristics through the internal resistance curve, record the change of the internal resistance value at each time point, and construct the internal resistance increase trend over time.

[0103] The Gaussian process regression (GPR) model is used to model the internal resistance offset characteristics, and the Gaussian process regression is used to predict the future offset of the internal resistance. The formula is as follows: ;in, For time Internal resistance value at the moment; is the mean function, which represents the average deviation of the internal resistance; is the covariance function, indicating different time and The correlation between represents another point in time, Together they are used to define correlations between different points in time; represents a Gaussian process.

[0104] Through Gaussian process regression modeling, future trend predictions of internal resistance shift can be generated, providing data support for further evaluation of granulation trends.

[0105] Set the internal resistance change threshold based on the normal fluctuation range in historical data. Set the internal resistance change threshold to ,when Beyond When , it indicates that the internal resistance fluctuation is abnormal, which may accelerate the granulation of active materials.

[0106] The granulation risk value is calculated to quantify the potential impact of internal resistance deviation on material granulation. The expression is: ;in, is the granular risk value, For the analysis period, Indicates the rate of change of internal resistance.

[0107] A larger particle size risk value means that the internal resistance fluctuation has a stronger potential impact on the particle size of the active material, which may accelerate the degradation of the material to a greater extent.

[0108] Set the granular risk threshold and compare the granular risk value with the granular risk threshold:

[0109] When the particle risk value is greater than the particle risk threshold, it is determined that the power battery internal resistance excursion has the risk of material degradation, indicating that the impact of the internal resistance fluctuation on the active material particle size exceeds the normal range and may cause material degradation. At this time, it is determined that the power battery internal resistance excursion has the risk of material degradation.

[0110] When the particle formation risk value is less than or equal to the particle formation risk threshold, it is determined that the internal resistance fluctuation of the power battery does not pose a risk to material degradation, which means that the impact of the internal resistance fluctuation on the active material is still within a safe range and has not reached the level of accelerating material degradation.

[0111] The particle risk threshold is set based on the historical data of the internal resistance fluctuation of the power battery under normal microcirculation, and the range of the average value and three times the standard deviation is determined through statistical analysis. The upper limit of the range is used as the particle risk threshold to ensure the distinction between normal fluctuations and abnormal deviations, so as to accurately determine the material particle risk.

[0112] Based on the potential risk of structural degradation caused by spectrum concentration anomaly and the potential impact of internal resistance fluctuation on active material granulation, it is determined whether the power battery has a risk of microcirculation anomaly, specifically:

[0113] When the potential risk of abnormal spectrum concentration to structural degradation is normal, and the fluctuation of internal resistance of the power battery does not pose a risk to material degradation, it is determined that there is no risk of abnormal microcirculation in the power battery; otherwise, it is determined that there is a risk of abnormal microcirculation in the power battery.

[0114] When the spectrum concentration abnormality and internal resistance fluctuation do not exceed the safety threshold, it indicates that the structural integrity and material stability of the power battery in the microcirculation state are within the normal range. The absence of significant abnormal characteristics means that the microcirculation has no significant impact on the physical structure and material properties of the battery, so it is determined that there is no risk of microcirculation abnormality in the power battery. If any factor exceeds the normal range, that is, the degree of spectrum concentration abnormality is high or the internal resistance fluctuation is significant, it means that the battery may be affected by stress concentration or material structure damage in the microcirculation state. This means that the microcirculation process may induce or aggravate structural degradation or granulation progression, increasing the risk of battery degradation. Therefore, it is determined that the power battery has a risk of microcirculation abnormality.

[0115] The stress accumulation tensor of the battery cell is analyzed by tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration. Specifically:

[0116] Collect stress data of power batteries under abnormal microcirculation conditions, and apply filtering methods to denoise and smooth the data:

[0117] In the abnormal state of microcirculation, the stress data of the power battery unit is collected through the sensor. This stress data reflects the stress of the power battery unit in different positions and directions, and the stress includes data in the three-dimensional directions of the x-axis, y-axis, and z-axis. The collected stress data may have certain noise interference due to the influence of external noise. Therefore, after the data collection is completed, data preprocessing is required to improve the data accuracy.

[0118] A low-pass filter is used to filter the collected stress data to eliminate high-frequency noise. The specific selection of the filtering method depends on the noise frequency range in the battery working environment, and a low-pass filter that meets the signal characteristics is usually selected.

[0119] The preprocessed stress data are organized according to three-dimensional spatial coordinates to construct a third-order stress tensor;

[0120] The processed stress data are organized according to three-dimensional spatial coordinates to construct a third-order stress tensor.

[0121] The stress data in these directions are integrated into a third-order tensor, expressed as: ;in, is the third-order stress tensor, , and Corresponding to the coordinate indexes on the x-axis, y-axis, and z-axis directions respectively, , as well as Represent the stress values ​​in the x-axis, y-axis and z-axis directions respectively.

[0122] The stress tensor is decomposed into tensor components in different directions to further reveal the local concentration of stress distribution in each direction:

[0123] Tensor decomposition is a technique that represents a tensor as a product of multiple low-order tensors or matrices, which can effectively separate the stress characteristics in each direction. Assume that the result of the third-order tensor decomposition is: ;in, is the third-order stress tensor before decomposition; is the tensor rank after tensor decomposition; Represents the order index of the tensor after decomposition, which is used to identify the stress characteristic components in different directions during the decomposition process; , and are the stress characteristic matrices in the x-axis, y-axis and z-axis directions respectively; Represents the tensor product operator, which is used in tensor decomposition to combine low-order tensors or matrices into high-order tensors.

[0124] By decomposing the tensor, the local stress distribution in each direction can be obtained. The components after tensor decomposition are used for further singular value decomposition to quantify the stress concentration.

[0125] Based on the results of tensor decomposition, the singular value decomposition method is used to extract the singular values ​​in the stress tensor:

[0126] Based on the tensor decomposition results, singular value decomposition (SVD) is used to decompose each feature matrix to extract the singular values. Singular value decomposition is a matrix decomposition technique that can decompose a matrix into multiple singular values ​​to help identify the main features in the matrix.

[0127] The singular value decomposition of the stress characteristic matrix is: ;in, represents the stress characteristic matrix; is the total number of singular values; Indicates singular values, representing the main eigenstrength of the matrix; and are the transposed forms of the left singular vector and the right singular vector corresponding to the singular values, respectively, and are used to form the decomposition of the matrix; is the number of the singular value; Represents the row index of the stress characteristic matrix, corresponding to a specific row of the matrix; Represents the column index of the stress characteristic matrix, corresponding to a specific column of the matrix; is the transpose symbol.

[0128] Through singular value decomposition, the maximum value among the singular values ​​can be obtained to quantify the stress concentration. The maximum value among the singular values ​​is selected as the main stress eigenvalue to calculate the stress damage influence index.

[0129] The stress damage influence index is calculated by the maximum value among the singular values ​​to quantify the degree of damage to the material structure caused by stress concentration:

[0130] The stress damage impact index is expressed as: ;in, is the stress damage impact index; is the maximum value among the singular values; It is a quantization function used to map the maximum singular value to a damage impact index. The form of the function depends on the battery material characteristics and experimental data, and can usually be a linear or nonlinear function.

[0131] The larger the stress damage impact index, the higher the degree of damage to the material structure caused by stress concentration. The stress damage impact index can be used as a risk parameter to assess the impact of stress on materials and provide support for subsequent battery safety management and maintenance decisions.

[0132] The attenuation gradient of the electrode polarization signal is analyzed by variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure caused by granulation acceleration. Specifically:

[0133] Collect electrode polarization signals when the microcirculation of the power battery is abnormal:

[0134] When the microcirculation of the power battery is abnormal, the sensor equipment is used to collect the electrode polarization signal (polarization signal data) of the battery electrode. The electrode polarization signal reflects the polarization behavior of the electrode under microcirculation pressure, which is manifested as the electrode polarization amplitude that changes with time.

[0135] The collected electrode polarization signal may fluctuate due to noise interference, so it needs to be preprocessed. To this end, a low-pass filter is used to filter out high-frequency noise from the signal to obtain a smooth electrode polarization signal.

[0136] The variational mode decomposition method is used to decompose the electrode polarization signal into different modes, and the frequency components and attenuation characteristics of each mode are obtained:

[0137] The preprocessed electrode polarization signal is subjected to variational mode decomposition to decompose the electrode polarization signal into several mode components with different frequency characteristics.

[0138] For each modal component, the frequency component and attenuation characteristics corresponding to the modal component are calculated.

[0139] The frequency component represents the main frequency of the mode, which is usually obtained through modal spectrum analysis; the attenuation feature represents the rate at which the amplitude of the modal signal decays over time. The frequency and attenuation feature pairs of each mode are obtained and used to construct the input data set of the recursive neural network.

[0140] A recursive neural network model is constructed based on the frequency components and attenuation characteristics of the modes to learn the granulation acceleration trends of different modes:

[0141] The input layer of the recurrent neural network receives the frequency components and attenuation characteristics of each mode, and identifies the changes in the attenuation gradient over time through a multi-layer network structure.

[0142] In the recurrent neural network model, the hidden layer parameters, weight parameters and bias parameters are used to define the activation function of the network as , the output of the recurrent neural network model is: .

[0143] Indicates at time Moment The network output result of the modality, this value represents the recurrent neural network model at time For The prediction result of the attenuation characteristics of each mode is the attenuation prediction value output by the model.

[0144] Represents the activation function, which defines the way of nonlinear transformation in the network. Common activation functions include ReLU (linear rectification), tanh (hyperbolic tangent), etc. The choice of activation function determines the nonlinear processing capability of the network.

[0145] Represents the hidden layer parameter matrix, which is used for weighting , to generate the hidden layer output of the neural network, and the values ​​of the hidden layer parameters are continuously updated during the model training process to optimize the prediction accuracy of the network.

[0146] Indicates The input feature vector of a mode contains the frequency components and attenuation characteristics of the mode. The input feature vector describes the basic characteristics of the modal signal and is the input of the recursive neural network.

[0147] Represents the weight matrix of the recurrent neural network. Used for the output of the previous time step The weights are applied to reflect the degree of influence of each step in the time series on subsequent predictions, and the weight matrix is ​​updated during the training process to adapt to the time dependency of the signal.

[0148] Indicates at time Moment The network output result of each mode represents the impact of the previous time step on the current time step. This value serves as feedback for the recurrent neural network to ensure that the network can remember the previous state and reflect the dynamic changes of the time series.

[0149] Represents the bias parameter, which is used to adjust the output of the activation function, improve the flexibility of the model, and ensure that the network can still produce non-zero output when there is no input signal. The bias parameter is updated during the model training process to optimize the overall performance of the network.

[0150] The recursive neural network model is trained to learn the attenuation gradients of different modes, thereby being able to capture the impact trend of accelerated granulation on the electrode structure.

[0151] The historical electrode polarization signals are imported into the recurrent neural network model for training, and the recurrent neural network model is used to predict the future changes of the attenuation gradient and obtain the granulation attenuation rate:

[0152] Import the historical electrode polarization signal data into the recursive neural network model and train the model. During the training process, optimize the model parameters , and This enables the network to more accurately predict the changing trend of the attenuation gradient over time. After training, the recursive neural network model is used to predict the future attenuation gradient. During the prediction process, the network gradually outputs the attenuation rate of each mode according to the time step, and finally obtains the granular attenuation rate of each mode.

[0153] Will be in time Moment The network output results of the modalities are compared with the Moment The difference between the network output results of the two modes is divided by the time interval between these two moments to obtain the granular decay rate.

[0154] The output of the recurrent neural network model is used as the prediction result of the attenuation gradient of each mode, and the particle cumulative degradation index is calculated to quantify the degree of degradation of the electrode structure due to particle acceleration:

[0155] The granulation decay rate of each mode is used to calculate the particle accumulation degradation index to quantify the influence of granulation on electrode structure degradation. The expression of particle accumulation degradation index is: ;in, is the particle cumulative degradation index, is the total number of modes, is the modality number, is the total time period for evaluation, It is The granular decay rate of each mode.

[0156] The higher the particle cumulative degradation index, the higher the degree of degradation of the electrode structure due to accelerated granulation, indicating that in multiple modes, the decay rate of the electrode polarization signal is large and a significant decay trend has accumulated over time. This rapid and significant decay reflects the accelerated granulation phenomenon inside the electrode material, that is, the microscopic particles of the material structure have become significantly loose or broken. This phenomenon will reduce the conductivity and mechanical strength of the electrode material, leading to structural degradation.

[0157] By comprehensively analyzing the degree of damage to the material structure caused by stress concentration and the degree of degradation of the electrode structure caused by accelerated granulation, the overall degradation trend of the power battery under microcirculation is evaluated; based on the overall degradation trend, the power battery management is optimized, specifically:

[0158] The stress damage influence index and the particle cumulative degradation index are normalized, and the normalized stress damage influence index and the particle cumulative degradation index are respectively assigned weight coefficients to calculate the overall degradation trend index. The calculation of the overall degradation trend index can be realized by the following formula, and its expression is: ;in, is the overall degradation trend index, and are the weight coefficients of stress damage influence index and particle cumulative degradation index, respectively, and and Both are greater than 0.

[0159] The larger the overall degradation trend index, the more obvious the overall degradation trend of the power battery in the microcirculation state, indicating that stress concentration damage and granular degradation are aggravated together, leading to an enhanced overall trend of structural degradation of battery materials. This means that battery materials are becoming increasingly fragile at the microscopic level, and the battery is more likely to experience problems such as capacity decay and shortened life during use. Therefore, when the overall degradation trend index is larger, the overall degradation risk of the power battery is higher, indicating that the battery is closer to a potential failure state.

[0160] The overall degradation trend threshold is set based on the historical data and experience data of the power battery. The specific method is to determine the reasonable range of stress concentration damage and granular degradation through a large amount of battery degradation trend data under normal and abnormal working conditions. Based on this, a threshold that meets the battery life and performance requirements is selected to distinguish between normal and significant degradation states.

[0161] Compare the overall degradation trend index to the overall degradation trend threshold:

[0162] When the overall degradation trend index is greater than the overall degradation trend threshold, it is determined that the overall degradation trend of the power battery is obvious. This indicates that the battery structure and materials are severely degraded, which may accelerate the battery capacity decay or cause failure. Preventive maintenance is recommended.

[0163] At this time, optimize the power battery management, specifically:

[0164] The battery management system (BMS) can appropriately reduce the charging current or limit the charging voltage to reduce the stress and granulation of battery materials. This dynamic adjustment can effectively slow down the degradation rate, reduce the pressure on the material structure under microcirculation conditions, and extend the overall life of the battery.

[0165] Increase the frequency of battery testing to more closely monitor the battery status. The battery management system can arrange proactive maintenance, such as equalizing battery cell voltages, detecting and replacing degraded battery cells, and preventing further expansion of degradation areas. These measures can identify potential problems in advance and deal with them in a timely manner to avoid more serious damage accumulation.

[0166] The battery management system can send out early warning signals to prompt users or maintenance personnel to perform preventive replacement or repair. This can avoid battery failure or failure due to uncontrolled degradation and reduce operational losses and safety risks caused by battery failure.

[0167] When the overall degradation trend index is less than or equal to the overall degradation trend threshold, the overall degradation trend of the power battery is judged to be normal. This indicates that the battery structure and material performance are stable, the degradation is within an acceptable range, and the battery can continue to operate normally.

[0168] Example 2

[0169] The difference between Example 2 of the present invention and Example 1 is that this example introduces a power battery management system based on data drive.

[0170] Figure 2 A structural schematic diagram of a data-driven power battery management system of the present invention is given. The data-driven power battery management system includes a spectrum clustering analysis module, an internal resistance shift analysis module, a cycle risk determination module, a stress accumulation analysis module, a polarization degradation analysis module and an overall degradation assessment module.

[0171] Spectrum clustering analysis module: Analyze the microcirculation energy spectrum distribution through the transient spectrum clustering algorithm to evaluate the potential risk of structural degradation caused by spectrum concentration anomalies.

[0172] Internal resistance drift analysis module: Analyze the drift trend of battery internal resistance and evaluate the potential impact of internal resistance fluctuations on active material granulation.

[0173] Cycle risk determination module: Based on the potential risk of structural degradation caused by spectrum concentration anomalies and the potential impact of internal resistance fluctuations on the granulation of active materials, it determines whether the power battery has microcirculation anomaly risks.

[0174] Stress accumulation analysis module: When there is a risk of microcirculation abnormality in the power battery, the stress accumulation tensor of the battery cell is analyzed through tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration.

[0175] Polarization degradation analysis module: When there is a risk of abnormal microcirculation in the power battery, the attenuation gradient of the electrode polarization signal is analyzed through variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure caused by granularization acceleration.

[0176] Overall degradation assessment module: By comprehensively analyzing the degree of damage to the material structure caused by stress concentration and the degree of degradation of the electrode structure caused by granulation acceleration, the overall degradation trend of the power battery under microcirculation conditions is assessed; and the power battery management is optimized based on the overall degradation trend.

[0177] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0178] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or may be transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0179] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0180] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0181] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0182] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0183] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0184] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for 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 methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0185] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0186] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A data-driven power battery management method, characterized in that: The steps include: The microcirculation energy spectrum distribution is analyzed by transient spectrum clustering algorithm to evaluate the potential risk of structural degradation caused by spectrum concentration anomalies. Analyze the deviation trend of battery internal resistance and evaluate the potential impact of internal resistance fluctuation on active material granulation; Based on the potential risk of structural degradation caused by spectrum concentration anomalies and the potential impact of internal resistance fluctuations on active material granulation, determine whether the power battery has a risk of microcirculation anomalies; When there is a risk of abnormal microcirculation in the power battery, the stress accumulation tensor of the battery cell is analyzed through tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration; the attenuation gradient of the electrode polarization signal is analyzed through variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure caused by granulation acceleration; By comprehensively analyzing the damage degree of stress concentration on material structure and the degradation degree of electrode structure accelerated by granulation, the overall degradation trend of power batteries under micro-circulation state is evaluated; Optimize power battery management based on this overall degradation trend; The microcirculation energy spectrum distribution is analyzed by transient spectrum clustering algorithm to evaluate the potential risk of structural degradation caused by spectrum concentration anomaly, specifically: Decompose the energy spectrum of the power battery in the microcirculation state and extract the amplitude and frequency characteristics of the spectrum; Divide the extracted amplitude features and frequency features into different frequency bands according to the frequencies, and generate characteristic data of the frequency bands; The characteristic data of each frequency band are hierarchically clustered by multi-scale decomposition method to analyze the fluctuation pattern and change rate of each frequency band; Calculate the mutation points and abnormal intervals in the spectrum and identify the spectrum energy deviation under the microcirculation state; According to the spectrum energy deviation and hierarchical clustering results, the potential risk of spectrum concentration anomaly to structural degradation is evaluated; The stress accumulation tensor of the battery cell is analyzed by tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration. Specifically: Collect stress data of power batteries under abnormal microcirculation conditions, and apply filtering methods to denoise and smooth the data; The preprocessed stress data are organized according to three-dimensional spatial coordinates to construct a third-order stress tensor; Perform tensor decomposition on the stress tensor to obtain stress components in different directions, revealing the local concentration of stress distribution in each direction; Based on the results of tensor decomposition, the singular value decomposition method is used to extract the singular values ​​in the stress tensor; The stress damage influence index is calculated by the maximum value among the singular values ​​to quantify the degree of damage to the material structure caused by stress concentration; The attenuation gradient of the electrode polarization signal is analyzed by variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure caused by granulation acceleration. Specifically: Collect electrode polarization signals when the microcirculation of the power battery is abnormal; The electrode polarization signal is decomposed into different modes using the variational mode decomposition method to obtain the frequency components and attenuation characteristics of each mode. A recursive neural network model is constructed based on the frequency components and attenuation characteristics of the modes to learn the granulation acceleration trends of different modes; The historical electrode polarization signals are imported into the recurrent neural network model for training, and the recurrent neural network model is used to predict the future changes of the attenuation gradient and obtain the granulation attenuation rate; The output of the recursive neural network model is used as the prediction result of the attenuation gradient of each mode, and the particle cumulative degradation index is calculated to quantify the degree of degradation of the electrode structure caused by particle acceleration.

2. The data-driven power battery management method according to claim 1, characterized in that: According to the spectrum energy offset and hierarchical clustering results, the potential risk of spectrum concentration anomaly to structural degradation is evaluated, specifically: Calculate the structural degradation risk value, the expression is: ;in, is the structural degradation risk value, represents the total number of clustering levels for hierarchical clustering, Indicates The weight factor of each clustering level, Indicates The fluctuation pattern of the cluster level, Indicates The rate of change of each cluster level; Set the structural degradation risk threshold and compare the structural degradation risk value with the structural degradation risk threshold: When the structural degradation risk value is greater than the structural degradation risk threshold, it is determined that the potential risk of the spectrum concentration anomaly to structural degradation is high; when the structural degradation risk value is less than or equal to the structural degradation risk threshold, it is determined that the potential risk of the spectrum concentration anomaly to structural degradation is normal.

3. The data-driven power battery management method according to claim 2, characterized in that: The deviation trend of the battery internal resistance is analyzed to evaluate the potential impact of internal resistance fluctuations on the granulation of active materials, specifically: Collect the internal resistance data of the power battery in the microcirculation state and extract the offset characteristics through the internal resistance curve; The Gaussian process regression model is used to model the internal resistance offset characteristics, and the Gaussian process regression is used to predict the future offset of the internal resistance. The formula is as follows: ;in, For time The internal resistance value at the moment, is the mean function, is the covariance function, Indicates another point in time. represents a Gaussian process; Calculate the granulation risk value, the expression is: ;in, is the granular risk value, For the analysis period, Indicates the rate of change of internal resistance; Set the granular risk threshold and compare the granular risk value with the granular risk threshold: When the particle risk value is greater than the particle risk threshold, it is determined that the internal resistance deviation of the power battery poses a risk of material degradation; when the particle risk value is less than or equal to the particle risk threshold, it is determined that the internal resistance fluctuation of the power battery does not pose a risk of material degradation.

4. The data-driven power battery management method according to claim 3 is characterized in that: Based on the potential risk of structural degradation caused by spectrum concentration anomalies and the potential impact of internal resistance fluctuations on active material granulation, it is determined whether the power battery has a risk of microcirculation anomalies, specifically: When the potential risk of structural degradation caused by the abnormal concentration of the spectrum is normal, and the fluctuation of the internal resistance of the power battery does not pose a risk of material degradation, it is determined that there is no risk of microcirculation abnormality in the power battery; otherwise, it is determined that there is a risk of microcirculation abnormality in the power battery.

5. The data-driven power battery management method according to claim 4, characterized in that: The particle cumulative degradation index is calculated to quantify the degree of degradation of the electrode structure due to particle acceleration, specifically: The granulation decay rate of each mode is used to calculate the particle cumulative degradation index, which is expressed as: ;in, is the particle cumulative degradation index, is the total number of modes, is the modality number, is the total time period for evaluation, It is The granular decay rate of each mode.

6. The data-driven power battery management method according to claim 5, characterized in that: By comprehensively analyzing the degree of damage to the material structure caused by stress concentration and the degree of degradation of the electrode structure caused by accelerated granulation, the overall degradation trend of the power battery under microcirculation is evaluated; based on the overall degradation trend, the power battery management is optimized, specifically: The stress damage influence index and the particle cumulative degradation index are normalized, and weight coefficients are assigned to the normalized stress damage influence index and the particle cumulative degradation index, respectively, to calculate the overall degradation trend index; Set the overall degradation trend threshold; compare the overall degradation trend index with the overall degradation trend threshold: When the overall degradation trend index is greater than the overall degradation trend threshold, it is determined that the overall degradation trend of the power battery is obvious; Optimize power battery management at this time; When the overall degradation trend index is less than or equal to the overall degradation trend threshold, it is determined that the overall degradation trend of the power battery is normal.

7. A data-driven power battery management system, used to implement the data-driven power battery management method according to any one of claims 1 to 6, characterized in that: It includes spectrum clustering analysis module, internal resistance shift analysis module, cycle risk determination module, stress accumulation analysis module, polarization degradation analysis module and overall degradation assessment module; Spectrum clustering analysis module: Analyze the microcirculation energy spectrum distribution through transient spectrum clustering algorithm to evaluate the potential risk of structural degradation caused by spectrum concentration anomaly; Internal resistance deviation analysis module: Analyze the deviation trend of battery internal resistance and evaluate the potential impact of internal resistance fluctuation on active material granulation; Cycle risk determination module: Based on the potential risk of structural degradation caused by spectrum concentration anomalies and the potential impact of internal resistance fluctuations on active material granulation, it determines whether the power battery has microcirculation anomaly risks; Stress accumulation analysis module: When there is a risk of abnormal microcirculation in the power battery, the stress accumulation tensor of the battery cell is analyzed through tensor decomposition and singular value decomposition to evaluate the degree of damage to the material structure caused by stress concentration; Polarization degradation analysis module: When there is a risk of abnormal microcirculation in the power battery, the attenuation gradient of the electrode polarization signal is analyzed through variational mode decomposition and recursive neural network to evaluate the degree of degradation of the electrode structure caused by granulation acceleration; Overall degradation assessment module: By comprehensively analyzing the damage degree of stress concentration to the material structure and the degradation degree of the electrode structure accelerated by granulation, the overall degradation trend of the power battery under microcirculation state is assessed; Optimize power battery management based on this overall degradation trend.

Images