Lithium battery health state assessment method and system

Abstract

The invention discloses a lithium battery health state assessment method and system, and particularly relates to the technical field of battery health state assessment. The method comprises the following steps: performing time sequence alignment and structured preprocessing on multi-source operation data of a target lithium battery in a plurality of historical work cycles to construct a structured data set; constructing a spatial-temporal characteristic residual error map based on residual error mapping analysis, and extracting a spatial heterogeneity index; in combination with a spatial heterogeneity index, generating regional degradation feature mapping; through high-dimensional feature embedding and evolution path clustering, a heterogeneous aging mode is identified, and a classification result is generated; evaluating the health state grade of the target lithium battery according to the regional degradation characteristic mapping and heterogeneous aging mode classification result; whether the battery has a local potential thermal runaway risk or not is judged based on the evaluation result, and a corresponding risk early warning signal and a safety disposal suggestion are generated, so that the nonlinear influence of the lithium battery aging heterogeneity can be accurately identified, and the health state evaluation precision and the safety risk early warning capability are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of battery state of health assessment, and more specifically, to a method and system for assessing the state of health of a lithium battery. Background Art

[0002] Traditional lithium battery state of health assessment technologies generally establish global degradation models based on the overall voltage-current curve, capacity retention rate, or single impedance characteristics, assuming that the internal aging process of the battery is uniform and the index evolution is monotonic. However, during battery cycling, the aging rates, aging modes, and aging degrees experienced by different regions inside the battery are inconsistent and non-uniform. Existing methods only infer the state of health from macroscopic average signals and cannot identify the masking effect of local premature degradation on the overall performance of the lithium battery, resulting in an amplified error in the remaining life prediction.

[0003] To solve the above problems, a technical solution is provided now. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method and system for assessing the state of health of a lithium battery to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions: A method for assessing the state of health of a lithium battery, comprising the following steps: S1: Collect multi-source operation data of the target lithium battery in multiple historical working cycles, perform time series alignment and structured preprocessing on the multi-source data based on a preset synchronization rule to obtain a structured data set; S2: Perform residual mapping analysis on the structured data set, construct a spatio-temporal feature residual map, and extract spatial heterogeneity indicators; S3: Identify the degradation trends of different regions inside the battery based on the spatial heterogeneity indicators, and generate a regional degradation feature map; S4: Perform high-dimensional feature embedding and evolution path clustering on the spatial heterogeneity indicators, identify the evolution paths of heterogeneous aging modes, and generate a classification result of heterogeneous aging modes; S5: Assess the state of health of the target lithium battery based on the regional degradation feature map and the classification result of heterogeneous aging modes, and output the state of health level of the target lithium battery; S6: According to the state of health level of the target lithium battery, determine whether there is a local potential thermal runaway risk for the target lithium battery, and output a warning signal and safety disposal suggestions.

[0006] In a preferred embodiment, S1 is specifically: Collect multi-source operation data of the target lithium battery in multiple historical working cycles; Synchronize the multi-source operation data according to the time stamp, and unify the time dimension using the preset time series alignment rules to obtain a multi-variable data set; Perform format standardization and missing value filling on the multi-variable data set to construct a structured data set.

[0007] In a preferred embodiment, S2 is specifically: Use the residual mapping analysis method to calculate the difference between each variable in the structured data set and the preset benchmark model; According to the preset time series window length, divide the difference data into multiple continuous time windows, and construct a feature residual matrix for each continuous time window; Perform two-dimensional spatial interpolation on each feature residual matrix to construct a spatio-temporal feature residual map; Extract the spatial heterogeneity index from the spatio-temporal feature residual map.

[0008] In a preferred embodiment, S3 is specifically: Based on the spatial heterogeneity index, divide the internal space of the target lithium battery into multiple independent analysis regions; According to the spatial heterogeneity index in each analysis region, fit the degradation trend curve of each analysis region over time, and extract the degradation trend characteristic parameters of each analysis region; Generate a regional degradation characteristic map according to the degradation trend characteristic parameters.

[0009] In a preferred embodiment, S4 is specifically: Use a kernel function to map the spatial heterogeneity index to a high-dimensional feature space to form a multi-dimensional feature vector; Apply the spectral clustering algorithm to the multi-dimensional feature vector to divide it into multiple heterogeneous aging mode categories; Sort the evolution paths of the order of each heterogeneous aging mode category in the time dimension to generate a heterogeneous aging mode classification result.

[0010] In a preferred embodiment, S5 is specifically: Input the regional degradation characteristic map and the heterogeneous aging mode classification result into the health state evaluation model to calculate the health state index of the target lithium battery; According to the preset health state level division standard, classify the health state index and output the health state level of the target lithium battery.

[0011] In a preferred embodiment, S6 is specifically: Combine the regional degradation characteristic map with the health state level of the target lithium battery to determine the comprehensive health index of each analysis region; For the comprehensive health index of each analysis area, determine whether there is a potential thermal runaway risk in the analysis area according to a preset risk threshold; For the analysis area determined to have a potential thermal runaway risk, output the corresponding local risk warning signal and safety disposal suggestions.

[0012] On the other hand, the present invention provides a lithium battery health state evaluation system, including: Data acquisition unit: Collect multi-source operation data of the target lithium battery in multiple historical working cycles, perform time series alignment and structured preprocessing on the multi-source data based on a preset synchronization rule to obtain a structured data set; Residual analysis unit: Perform residual mapping analysis on the structured data set, construct a spatio-temporal feature residual map, and extract spatial heterogeneity indicators; Trend modeling unit: Identify the degradation trends of different regions inside the battery based on the spatial heterogeneity indicators, and generate a regional degradation feature map; Clustering recognition unit: Perform high-dimensional feature embedding and evolutionary path clustering on the spatial heterogeneity indicators, identify the evolutionary paths of heterogeneous aging patterns, and generate a heterogeneous aging pattern classification result; Health assessment unit: Evaluate the health state of the target lithium battery based on the regional degradation feature map and the heterogeneous aging pattern classification result, and output the health state level of the target lithium battery; Risk warning unit: Judge whether there is a local potential thermal runaway risk in the target lithium battery according to the health state level of the target lithium battery, and output a warning signal and safety disposal suggestions.

[0013] Technical effects and advantages of the lithium battery health state evaluation method and system of the present invention: By performing time series synchronization and structured preprocessing on multi-source operation data, the integrity and consistency of the data are realized; based on residual mapping, a spatio-temporal feature residual map is constructed to accurately reveal the regional heterogeneity inside the battery; combined with spatial heterogeneity indicators, the degradation trends of each region are identified to accurately depict local degradation characteristics; high-dimensional feature embedding and evolutionary path clustering are used to identify the temporal evolution characteristics of heterogeneous aging patterns; taking the regional feature map and heterogeneous aging pattern classification result as inputs, the health state level is output through a health state evaluation model; based on the health level, the local thermal runaway risk is accurately judged and a warning signal and safety disposal suggestions are issued, realizing refined prevention and control of local fault risks, and significantly enhancing the operation safety and reliability of the battery. Description of the drawings

[0014] Figure 1 It is a schematic diagram of a lithium battery health state evaluation method of the present invention; Figure 2 It is a structural schematic diagram of a lithium battery health state evaluation system of the present invention. Detailed implementation manners

[0015] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0016] Embodiment 1 Figure 1 A method for evaluating the health state of a lithium battery according to the present invention is provided, which includes the following steps: S1: Collect multi-source operation data of the target lithium battery in multiple historical working cycles, perform time series alignment and structured preprocessing on the multi-source data based on a preset synchronization rule, and obtain a structured data set; S2: Perform residual mapping analysis on the structured data set, construct a spatio-temporal feature residual map, and extract spatial heterogeneity indicators; S3: Identify the degradation trends of different regions inside the battery based on the spatial heterogeneity indicators, and generate a regional degradation feature map; S4: Perform high-dimensional feature embedding and evolutionary path clustering on the spatial heterogeneity indicators, identify the evolutionary paths of heterogeneous aging patterns, and generate a classification result of heterogeneous aging patterns; S5: Evaluate the health state of the target lithium battery based on the regional degradation feature map and the classification result of heterogeneous aging patterns, and output the health state level of the target lithium battery; S6: According to the health state level of the target lithium battery, judge whether there is a local potential thermal runaway risk for the target lithium battery, and output a warning signal and safety disposal suggestions.

[0017] S1: Collect multi-source operation data of the target lithium battery in multiple historical working cycles, perform time series alignment and structured preprocessing on the multi-source data based on a preset synchronization rule, and obtain a structured data set, including: Collect multi-source operation data of the target lithium battery in multiple historical working cycles; Specifically, set the continuous operation cycle of the lithium battery in the battery test platform. Each working cycle includes a complete charging stage, a constant voltage holding stage, a standing stage, and a discharging stage. Through a multi-channel data acquisition device, the following operating data is continuously acquired in each working cycle: Voltage data is acquired through a high-precision voltage sensor connected to the positive and negative terminals of the battery; current data is acquired in real time through a Hall current sensor connected in series in the main circuit of the battery; electrochemical impedance spectroscopy data is obtained by connecting an electrochemical impedance analyzer to the battery terminals, applying an alternating excitation signal and measuring the response signal at multiple set frequency points; battery surface temperature data is acquired through thermocouple sensors arranged at different positions on the surface of the battery case to ensure comprehensive coverage of the temperature information of different spatial positions of the target lithium battery.

[0018] During the acquisition process, set the sampling frequency to ensure that the sampling rates of the voltage data, current data, and battery surface temperature data are consistent, and the sampling frequency of the electrochemical impedance spectroscopy data is determined according to the actual working conditions of the target lithium battery.

[0019] Synchronize the multi-source operating data according to the time stamps, and use the preset time sequence alignment rules to unify the time dimension to obtain a multi-variable data set; Specifically, record the corresponding accurate time stamp information during the acquisition of multi-source operating data, and calibrate it using a unified time server to ensure the accuracy and consistency of the time stamps recorded by each sensor; then use the preset time sequence alignment rules to perform unified time dimension alignment processing on the voltage data, current data, electrochemical impedance spectroscopy data, and battery surface temperature data, that is, perform interpolation or truncation operations on the missing or misaligned data in multiple data sequences. For example, when the time stamps of the current data do not match those of the voltage data, the linear interpolation method is used to obtain current data points consistent with the corresponding moments of the voltage data between two adjacent current data points, so that each item of multi-source operating data corresponds consistently in the same time dimension, forming a multi-variable data set on a unified time scale.

[0020] Perform format standardization and missing value filling on the multi-variable data set to construct a structured data set; Specifically, perform data format standardization on the multi-variable data set, that is, convert the voltage data, current data, electrochemical impedance spectroscopy data, and battery surface temperature data into a unified data format structure respectively, for example, all are converted into floating-point data formats, and at the same time unify the precision requirements of each data, such as the voltage precision to the millivolt level, the current precision to the milliampere level, and the temperature precision to two decimal places after the degree Celsius, to achieve unified data analysis precision.

[0021] Complete the filling process for missing data points in the multi-variable data set. Specifically, use the moving window average method of the same variable data within the adjacent time window for filling: take the arithmetic average of multiple known data points before and after the missing position to obtain the estimated value of the missing data, and fill it into the corresponding position to ensure the continuity and integrity of the entire multi-variable data set without any missing data points.

[0022] After completing the above standardization and missing value filling processes, construct a structured data set that includes voltage sequences, current sequences, electrochemical impedance spectroscopy sequences, and battery surface temperature sequences. The specific form of the structured data set is a two-dimensional data table with multiple rows and columns. Each row represents a unified sampling moment, and each column represents different types of data variables.

[0023] S2: Conduct residual mapping analysis on the structured data set, construct a spatio-temporal feature residual map, and extract spatial heterogeneity indicators, including: Calculate the difference between each variable in the structured data set and a preset reference model using the residual mapping analysis method; Specifically, the construction method of the preset reference model is to use the operating data of the target lithium battery in the initial healthy state as the reference state data. Perform point-by-point difference calculations on the voltage sequences, current sequences, electrochemical impedance spectroscopy sequences, and battery surface temperature sequences collected from the target lithium battery in different historical operating cycles with the corresponding variable sequences in the reference state data. After subtracting the data at the corresponding positions of each variable, obtain the difference data sequence of the corresponding variable to represent the degree of change of each variable in the current state of the target lithium battery relative to the initial healthy state. The difference calculation process is as follows: take the data value of a certain variable in the structured data set at a specific moment, subtract the data value of the corresponding variable in the reference model at the same moment, calculate the residual value of a certain variable at the specific moment, and repeat this calculation until all the data in the structured data set at all moments are processed, and finally obtain a complete difference data set.

[0024] Divide the difference data into multiple consecutive time windows according to the preset time series window length, and construct a feature residual matrix for each consecutive time window respectively; Specifically, divide the difference data of the historical operating cycle into multiple non-overlapping and equally long consecutive time windows. Each consecutive time window contains the difference data of multiple consecutive sampling moments to ensure that the data volume covered by each time window is the same and the time span is the same. The determination basis for the length of the consecutive time window is to ensure that the difference data within each time window can fully represent the electrochemical characteristic changes of the target lithium battery within the corresponding time window. The specific window length is determined according to the actual operating characteristics and data sampling frequency of the target lithium battery. For example, each consecutive time window can cover the data of a complete charge-discharge process to ensure the integrity and representativeness of the difference data within the time window.

[0025] The method for constructing the feature residual matrix is as follows: taking each sampling moment within each continuous time window as the row coordinates of the matrix, and each variable in the structured dataset as the column coordinates of the matrix. The value of each matrix element is the difference calculated between the variable at that sampling moment and the preset reference model, so as to obtain a feature residual matrix that can characterize the changing trends of multiple variables over time within each continuous time window.

[0026] Perform two-dimensional spatial interpolation processing on each feature residual matrix to construct a spatio-temporal feature residual map; Specifically, the interpolation method selects the Kriging interpolation method. Through the Kriging interpolation method, the covariance relationship between spatial positions can be fully considered for the estimation of differences. Specifically: set the coordinate grid system of the internal space of the lithium battery, and the coordinates of each grid node in the coordinate grid system represent specific spatial positions within the lithium battery. For each element in the feature residual matrix, use the difference data at multiple known measurement spatial positions closest to the grid node to calculate the estimated difference at the position to be interpolated through spatial correlation. The specific calculation process is: apply weights to multiple known difference data closest to the grid node and calculate the weighted sum. The calculated weights are determined according to the covariance function between spatial positions; finally, complete continuous interpolation data is obtained at the grid node positions within the battery internal space, thereby completing the two-dimensional spatial interpolation processing of each feature residual matrix and obtaining a complete spatio-temporal feature residual map that can reflect the changing laws of spatial positions and time dimensions.

[0027] Extract the spatial heterogeneity index from the spatio-temporal feature residual map; Specifically, for each grid node position in the spatio-temporal feature residual map, set the observation period in the time dimension as a time series interval of a fixed length, and extract the time series of the interpolation data of the grid node within the observation period. For the time series, calculate the corresponding variance value as the time change amplitude index of the grid node. The variance calculation formula used is: sum the squared differences between each interpolation at each moment in the time series and its average value, and then divide by the series length to obtain the spatial heterogeneity index.

[0028] S3: Identify the degradation trends of different regions inside the battery based on the spatial heterogeneity index, and generate a regional degradation feature map, including: Based on the spatial heterogeneity index, divide the internal space of the target lithium battery into multiple independent analysis regions; Specifically, the area division method is as follows: according to the numerical distribution of the spatial heterogeneity index at different positions inside the battery, the density clustering analysis method is used to cluster and divide each spatial position inside the battery. The density clustering analysis method is to classify and aggregate spatial positions according to the density degree of the spatial heterogeneity index in space and the similarity of spatial distances; the key parameters for setting the density clustering analysis method include the spatial distance threshold and the minimum density threshold. The spatial distance threshold is determined based on the internal structure size of the battery, and the minimum density threshold is determined by analyzing the distribution characteristics of the number of samples in the neighborhood of each grid node according to the local point density distribution of the spatial heterogeneity index at the grid positions inside the battery; through the above clustering and division method, the internal space of the target lithium battery is divided into multiple independent and clearly spatially distributed analysis areas. Each analysis area contains several spatial grid nodes, and the spatial positions between the analysis areas do not overlap with each other, providing a basis for the analysis of the area degradation trend.

[0029] According to the spatial heterogeneity index in each analysis area, the degradation trend curve of each analysis area changing with time is fitted, and the degradation trend characteristic parameters of each analysis area are extracted; Specifically, taking the spatial heterogeneity index of all spatial grid nodes in each analysis area as sample points, taking time as the independent variable and the spatial heterogeneity index as the dependent variable, a non-linear trend fitting model is selected for fitting the degradation trend curve. The non-linear trend fitting model is specifically an exponential decay fitting model; the exponential decay fitting model realizes parameter estimation through the least squares method. The fitting process is: taking the sum of the squares of the errors between the measured values of the spatial heterogeneity index in the area and the predicted values of the exponential decay fitting model as the minimum as the goal, calculating the parameters of the exponential decay fitting model to make the output result of the exponential decay fitting model closest to the actual measurement data as much as possible; finally, the degradation trend curve in each analysis area is obtained.

[0030] The degradation trend characteristic parameters specifically include the trend decay rate parameter and the decay amplitude parameter. The trend decay rate parameter is obtained by calculating the derivative of the area degradation trend curve, specifically the slope value of the fitted trend curve at different times, and the degradation rate is quantified by calculating the average value of the slope values; the decay amplitude parameter is obtained by calculating the numerical difference between the initial value and the termination value of the fitted trend curve to quantify the overall deterioration degree of each area during the entire observation period. Through the above method, the degradation trend curves in each analysis area can be quantified into characteristic parameters such as the trend decay rate and the decay amplitude, realizing the quantitative comparison of each analysis area.

[0031] According to the degradation trend characteristic parameters, a regional deterioration characteristic map is generated; Specifically, the construction process of the regional degradation feature mapping is as follows: using the trend decay rate parameter and the decay amplitude parameter as the coordinate axes of two dimensions, a mapping coordinate system is established in the two-dimensional space, and the trend decay rate parameter and the decay amplitude parameter of each analysis region are marked as the coordinate points in the mapping coordinate system. The regional degradation feature mapping is a mapping coordinate system composed of multiple analysis regions in the degradation trend feature parameters (trend decay rate parameter and trend decay amplitude parameter). The regional degradation feature mapping reveals the differential relationship in the spatial degradation degree among the analysis regions, providing an accurate spatial basis for the regional heterogeneity analysis and the regional positioning of the thermal runaway risk in the lithium battery health state assessment.

[0032] S4: Perform high-dimensional feature embedding and evolutionary path clustering on the spatial heterogeneity index to identify the evolutionary paths of heterogeneous aging patterns and generate the classification results of heterogeneous aging patterns, including: Use a kernel function to map the spatial heterogeneity index to a high-dimensional feature space to form a multi-dimensional feature vector; Specifically, define the distance measurement method between data points in the kernel function as the Euclidean distance, that is, the distance between any two spatial heterogeneity index data points is obtained by calculating the square root of the sum of the squares of the corresponding dimension numerical differences of the two spatial heterogeneity index data points; according to the calculation rules of the selected kernel function, perform an exponential decay transformation on the Euclidean distance between any two spatial heterogeneity index data points, that is, after taking the opposite of the square of the Euclidean distance value, and then through exponential function operation, obtain the output value of the kernel function; through the above kernel mapping calculation, the spatial heterogeneity index at each spatial position is transformed into a feature vector in the high-dimensional space, and a multi-dimensional feature vector set is obtained.

[0033] Apply the spectral clustering algorithm to the multi-dimensional feature vector to divide into multiple heterogeneous aging pattern categories; Specifically, all multi-dimensional feature vectors in the high-dimensional feature space are paired pairwise, and a feature similarity matrix is constructed by calculating the cosine similarity between each pair of feature vectors. A Laplacian matrix required for the spectral clustering algorithm is constructed using the similarity matrix. Specifically, the Laplacian matrix is obtained by transforming the similarity matrix into the difference between the degree matrix and the similarity matrix. The Laplacian matrix is subjected to eigen-decomposition operation to obtain an eigenvector matrix. According to the distribution of the eigenvector elements of the eigenvector matrix, a clustering method is used to cluster and divide the spatial heterogeneity index to obtain multiple heterogeneous aging mode categories. The clustering method uses the classical K-means clustering method, and the number of clusters is determined according to the clustering effect evaluation index. The clustering effect evaluation index is the silhouette coefficient of the clustering result, and the number of clusters that makes the silhouette coefficient reach the optimal value is selected as the basis for determining the number of categories. A heterogeneous aging mode refers to the local aging trend of a lithium battery during the aging process, which is different from the general law of the overall lithium battery due to various factors such as internal structure differences, chemical composition inhomogeneity, and temperature field inconsistency; for example: obvious differences in the aging rate (the capacity attenuation or internal resistance growth rate in a local area is significantly faster than other areas); differences in the severity of aging (some areas show early degradation characteristics, while other areas do not show obvious degradation). A heterogeneous aging mode category refers to a heterogeneous aging mode with a common aging characteristic trend; for example: Category 1 (rapid capacity decay category): dominated by rapid capacity decay, showing a high trend decay rate; Category 2 (temperature-sensitive category): the aging speed varies significantly with local temperature, showing temperature-dependent aging.

[0034] Sort the evolution paths of the different heterogeneous aging mode categories in the time dimension to generate the classification result of the heterogeneous aging mode. Specifically, for each heterogeneous aging mode category, calculate the average occurrence time of all multi-dimensional feature vectors within the heterogeneous aging mode category in the time dimension, that is, the category average time position, which is achieved by taking the arithmetic mean of the timestamp data corresponding to all multi-dimensional feature vectors within the category. Sort the categories according to the calculated category average time position, and the sorting method is the ascending order of the category average time position values from small to large, so as to reflect the evolution order of different heterogeneous aging mode categories appearing in sequence during the battery operation cycle. Through the above sorting, a clear and distinct classification result of the heterogeneous aging mode evolution path is obtained, providing data input for the overall health state level assessment of the lithium battery.

[0035] S5: Based on the regional degradation feature mapping and the classification result of the heterogeneous aging mode, evaluate the health state of the target lithium battery and output the health state level of the target lithium battery, including: Input the regional degradation feature mapping and the classification result of the heterogeneous aging mode into the health state evaluation model to calculate the health state index of the target lithium battery. Specifically, the health status assessment model is specifically a multi-parameter fusion assessment model. The specific implementation method of the multi-parameter fusion assessment model is as follows: The degradation trend characteristic parameters in the regional degradation characteristic mapping of each analysis area, including the trend decay rate parameter and the decay amplitude parameter, and the heterogeneous aging mode category characteristics corresponding to the analysis area are used as inputs at the same time. The specific calculation method of the multi-parameter fusion assessment model is to perform a linear combination of the trend decay rate parameter and the decay amplitude parameter of each analysis area to obtain a preliminary regional degradation index; perform a non-linear fusion calculation on the preliminary regional degradation index and the heterogeneous aging mode category characteristics. The non-linear fusion calculation method uses a feed-forward neural network, configured as a three-layer network structure, including an input layer, a hidden layer, and an output layer; the number of neurons in the input layer corresponds to the dimension of the input parameters, and the number of neurons in the hidden layer is determined through repeated experiments and optimized with the goal of minimizing the output error of the multi-parameter fusion assessment model; the single output value of the neurons in the output layer is the health status index of the target lithium battery; the training process of the feed-forward neural network uses a supervised learning method, and the supervised learning method is the backpropagation algorithm; the implementation of the backpropagation algorithm is to use the historical known actual battery aging data as training data, and take the minimum of the sum of the squares of the errors between the actual data and the output results of the multi-parameter fusion assessment model as the optimization goal, and repeatedly adjust the network weight parameters until the error converges; after the above process is trained, the regional degradation characteristic mapping and the heterogeneous aging mode classification results are input into the trained feed-forward neural network to calculate the health status index of the target lithium battery.

[0036] According to the preset health status level division standard, classify the health status index to output the health status level of the target lithium battery; Specifically, the health status level division standard is the numerical interval level division method: preset the first threshold and the second threshold; compare the health status index with the preset first threshold and second threshold. When the health status index is less than the first threshold, it is judged that the health status level of the target lithium battery is severely degraded; when the health status index is greater than or equal to the first threshold and less than the second threshold, it is judged that the health status level of the target lithium battery is slightly degraded; when the health status index is greater than or equal to the second threshold, it is judged that the health status level of the target lithium battery is in good health.

[0037] S6: According to the health status level of the target lithium battery, judge whether the target lithium battery has a local potential thermal runaway risk, and output a warning signal and safety disposal suggestions, including: Combine the regional degradation characteristic mapping with the health status level of the target lithium battery to determine the comprehensive health index of each analysis area; Specifically, the trend decay rate parameter and the trend decay amplitude parameter mapped by the above regional degradation characteristics are associated and fused with the overall health state level of the lithium battery, and an additive fusion method is adopted: different fusion weights are assigned to the trend decay rate parameter, the trend decay amplitude parameter, and the overall health state level of the lithium battery. The fusion weights are determined through statistical regression analysis of the measured data of historical lithium batteries. Specifically, the fusion weights are determined with the optimization of the statistical regression analysis results of historical data as the goal, so as to make the fusion effect of the regional characteristic parameters and the overall health state level reach the optimal; the fused value is the comprehensive health index corresponding to each analysis area, thereby reflecting the correlation between the regional characteristics and the overall lithium battery state level.

[0038] For the comprehensive health index of each analysis area, judge whether there is a potential thermal runaway risk in the analysis area according to the preset risk threshold; Specifically, the preset risk threshold is determined by statistically analyzing the historical operation data of the lithium battery to determine the critical value that can effectively distinguish the safe area from the risk area, which is the risk critical point of the comprehensive health index obtained through historical data analysis; the judgment process is to compare the comprehensive health index of each analysis area with the risk threshold one by one. If the value of the comprehensive health index of the analysis area is less than the preset risk threshold, it is judged that the analysis area has a potential thermal runaway risk; if the comprehensive health index of the analysis area is greater than or equal to the risk threshold, it is judged that the analysis area does not have a potential thermal runaway risk.

[0039] For the analysis area determined to have a potential thermal runaway risk, output the corresponding local risk warning signal and safety disposal suggestions; Specifically, generate a local risk warning signal for the analysis area determined to have a potential thermal runaway risk. The specific warning signal is a comprehensive warning prompt information including clear spatial position information and real-time monitoring suggestions; the spatial position information specifically includes the position coordinates of the risk area in the internal coordinate system of the lithium battery; the real-time monitoring suggestions are the continuous monitoring methods given for the risk area, including, for example, increasing the monitoring frequency and increasing the sensor layout density, etc., to ensure the safety monitoring intensity of the risk area.

[0040] Generate safety disposal suggestions to achieve the safety guarantee of the lithium battery operation. The safety disposal suggestions specifically include: implementing safety disposal schemes such as load adjustment, power limit, local temperature control strengthening, charge and discharge rate adjustment, etc. for the risk area; for example, for the analysis area with a potential thermal runaway risk, give suggestions to reduce the charge and discharge rate and improve the cooling capacity, so that the reduction effect of the disposal scheme on the risk of the analysis area reaches the optimal.

[0041] Embodiment 2 The difference between Embodiment 2 and Embodiment 1 of the present invention is that this embodiment introduces a lithium battery health state evaluation system.

[0042] Figure 2 The structural schematic diagram of a lithium battery health state evaluation system according to the present invention is given. A lithium battery health state evaluation system includes: Data acquisition unit: Collect multi-source operation data of the target lithium battery in multiple historical working cycles, perform time series alignment and structured preprocessing on the multi-source data based on a preset synchronization rule to obtain a structured data set; Residual analysis unit: Perform residual mapping analysis on the structured data set, construct a spatio-temporal feature residual map, and extract spatial heterogeneity indexes; Trend modeling unit: Identify the degradation trends of different regions inside the battery based on the spatial heterogeneity indexes, and generate a regional degradation feature map; Clustering recognition unit: Perform high-dimensional feature embedding and evolution path clustering on the spatial heterogeneity indexes, identify the evolution paths of heterogeneous aging patterns, and generate a classification result of heterogeneous aging patterns; Health assessment unit: Evaluate the health state of the target lithium battery based on the regional degradation feature map and the classification result of heterogeneous aging patterns, and output the health state level of the target lithium battery; Risk warning unit: According to the health state level of the target lithium battery, judge whether there is a local potential thermal runaway risk for the target lithium battery, and output a warning signal and safety disposal suggestions.

[0043] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0044] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can 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 processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0045] Those of ordinary skill in the art will realize that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0046] 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 foregoing method embodiments and will not be described herein again.

[0047] 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 merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, 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 displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or modules can be in electrical, mechanical, or other forms.

[0048] The module described as a separation component may or may not be physically separated. The component presented as a module may or may not be a physical module, and it may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0049] In addition, in each embodiment of this application, the various functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0050] If the described function is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0051] As described above, this is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claimed rights.

[0052] Finally: The above description is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for evaluating the health state of a lithium battery, characterized in that, It includes the following steps: S1: Collect multi-source operation data of the target lithium battery in multiple historical working cycles, perform time-series alignment and structured preprocessing on the multi-source data based on a preset synchronization rule to obtain a structured data set; S2: Perform residual mapping analysis on the structured data set, construct a spatio-temporal feature residual map, and extract spatial heterogeneity indicators; S3: Identify the degradation trends of different regions inside the battery based on the spatial heterogeneity indicators, and generate a regional degradation feature map; S4: Perform high-dimensional feature embedding and evolutionary path clustering on the spatial heterogeneity indicators, identify the evolutionary paths of heterogeneous aging patterns, and generate a classification result of heterogeneous aging patterns; S5: Evaluate the health status of the target lithium battery based on the regional degradation feature map and the classification result of heterogeneous aging patterns, and output the health status level of the target lithium battery; S6: According to the health status level of the target lithium battery, determine whether there is a local potential thermal runaway risk for the target lithium battery, and output a warning signal and safety disposal suggestions.

2. The method for evaluating the health state of a lithium battery according to claim 1, wherein S1 specifically is: Collect multi-source operation data of the target lithium battery in multiple historical working cycles; Synchronize the multi-source operation data according to the time stamps, and unify the time dimension using the preset time-series alignment rule to obtain a multi-variable data set; Perform format standardization and missing value filling on the multi-variable data set to construct a structured data set.

3. The method for evaluating the health state of a lithium battery according to claim 2, characterized in that, S2 specifically is: Use the residual mapping analysis method to calculate the difference between each variable in the structured data set and a preset benchmark model; Divide the difference data into multiple consecutive time windows according to the preset time-series window length, and construct a feature residual matrix for each consecutive time window; Perform two-dimensional spatial interpolation processing on each feature residual matrix to construct a spatio-temporal feature residual map; Extract spatial heterogeneity indicators from the spatio-temporal feature residual map.

4. The method for evaluating the health state of a lithium battery according to claim 3, characterized in that S3 specifically is: Based on the spatial heterogeneity indicators, divide the internal space of the target lithium battery into multiple independent analysis regions; According to the spatial heterogeneity indicators in each analysis region, fit the degradation trend curve of each analysis region over time, and extract the degradation trend characteristic parameters of each analysis region; Generate a regional degradation feature map according to the degradation trend characteristic parameters.

5. The method for evaluating the health state of a lithium battery according to claim 4, wherein S4 specifically is: Use a kernel function to map the spatial heterogeneity indicators to a high-dimensional feature space to form a multi-dimensional feature vector; Apply a spectral clustering algorithm to the multi-dimensional feature vector to divide it into multiple heterogeneous aging pattern categories; Sort the evolutionary paths of the heterogeneous aging pattern categories in the time dimension to generate a classification result of heterogeneous aging patterns.

6. The method for evaluating the health state of a lithium battery according to claim 5, wherein, S5 specifically is: Input the regional degradation feature map and the classification result of heterogeneous aging patterns into a health status evaluation model to calculate the health status index of the target lithium battery; According to the preset health status level division standard, perform grading processing on the health status index and output the health status level of the target lithium battery.

7. The method for evaluating the health state of a lithium battery according to claim 6, wherein, S6 specifically is: Combine the regional degradation feature map with the health status level of the target lithium battery to determine the comprehensive health index of each analysis region; For the comprehensive health index of each analysis region, judge whether there is a potential thermal runaway risk in the analysis region according to the preset risk threshold; For the analysis area where potential thermal runaway risks are determined, corresponding local risk warning signals and safety disposal suggestions are output.

8. A lithium battery state of health assessment system for implementing the lithium battery state of health assessment method according to any one of claims 1-7, characterized in that, Including: Data acquisition unit: Collect multi-source operation data of the target lithium battery in multiple historical working cycles, perform time series alignment and structured preprocessing on the multi-source data based on preset synchronization rules, and obtain a structured data set; Residual analysis unit: Perform residual mapping analysis on the structured data set, construct a spatio-temporal feature residual map, and extract spatial heterogeneity indicators; Trend modeling unit: Identify the deterioration trends of different regions inside the battery based on the spatial heterogeneity indicators, and generate a regional deterioration feature map; Clustering identification unit: Perform high-dimensional feature embedding and evolutionary path clustering on the spatial heterogeneity indicators, identify the evolutionary paths of heterogeneous aging patterns, and generate classification results of heterogeneous aging patterns; Health assessment unit: Evaluate the health status of the target lithium battery based on the regional deterioration feature map and the classification results of heterogeneous aging patterns, and output the health status level of the target lithium battery; Risk warning unit: According to the health status level of the target lithium battery, judge whether there is a local potential thermal runaway risk for the target lithium battery, and output warning signals and safety disposal suggestions.

Images