Battery replacement cabinet battery aging evaluation method and system based on charging mode clustering

Abstract

The invention discloses a battery changing cabinet battery aging evaluation method and system based on charging mode clustering, and relates to the technical field of battery changing cabinet battery detection. The method comprises the following steps: analyzing a battery charging cycle according to a charging period of a battery changing cabinet, and constructing a multi-dimensional charging characteristic; clustering learning is carried out based on the features, a plurality of charging mode clusters are constructed, health analysis is carried out on the battery based on the clusters, and battery health state parameters are obtained; calling real-time charging data, matching the charging mode cluster in combination with the health state parameters, and determining a target charging mode cluster; and performing battery aging evaluation according to the target charging mode cluster to generate an aging evaluation result. The technical problems of low evaluation precision and poor real-time performance caused by charging mode diversity of an existing battery aging evaluation method are solved, and the technical effects that refined analysis and dynamic matching of the battery health state are achieved based on charging mode clustering, and the accuracy, scene adaptability and real-time responsiveness of a battery aging evaluation result are improved are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery detection of battery swap cabinets, and particularly relates to a battery aging evaluation method and system for battery swap cabinets based on charging mode clustering. BACKGROUND

[0002] With the popularity of electric vehicles, battery swap cabinets, as important facilities for battery exchange and charging, are increasingly becoming an important part of electric vehicle charging infrastructure. Battery swap cabinets provide a more efficient battery replacement method for electric vehicles, which can shorten the battery charging time and improve the charging efficiency compared to traditional charging piles. However, the health status of the batteries in the battery swap cabinet directly affects the performance and user experience of electric vehicles. Battery aging is an inevitable problem during the use of the battery, and excessive aging not only affects the battery's endurance, but also may cause safety hazards.

[0003] Existing battery aging evaluation methods usually rely on simple battery capacity decay measurements, lack comprehensive analysis of batteries under different charging modes, and do not fully consider the complex working environment and charging conditions faced by batteries in actual use. This method may not accurately reflect the true aging of the battery, resulting in insufficient accuracy and reliability of the evaluation results. Therefore, how to comprehensively and accurately evaluate the aging of the battery based on the charging mode has become an important technical challenge in battery management and maintenance. SUMMARY

[0004] The present application provides a battery aging evaluation method and system for battery swap cabinets based on charging mode clustering, which solves the technical problem of low evaluation accuracy and poor real-time performance caused by the diversity of charging modes in existing battery aging evaluation methods.

[0005] In a first aspect, the present application provides a battery aging evaluation method for battery swap cabinets based on charging mode clustering, which comprises: carrying out charging cycle analysis on the battery of the battery swap cabinet according to the charging period of the battery swap cabinet, constructing multi-dimensional charging features; based on the multi-dimensional charging features, carrying out clustering learning to construct multiple charging mode clusters, based on the multiple charging mode clusters, carrying out battery health analysis on the battery of the battery swap cabinet to obtain battery health status parameters; retrieving real-time charging data of the battery of the battery swap cabinet in combination with the battery health status parameters to determine the target charging mode cluster; and carrying out battery aging evaluation on the battery of the battery swap cabinet according to the target charging mode cluster to generate a battery aging evaluation result.

[0006] In a second aspect, the present application provides a battery aging evaluation system for battery swap cabinets based on charging mode clustering, which comprises: The charging cycle analysis module: the battery of the battery swap cabinet is analyzed according to the charging period of the battery swap cabinet, and multi-dimensional charging characteristics are constructed; the battery health analysis module: based on the multi-dimensional charging characteristics, a plurality of charging mode clusters are constructed through clustering learning, and the battery health of the battery of the battery swap cabinet is analyzed based on the plurality of charging mode clusters to obtain battery health state parameters; the matching determination module: the real-time charging data of the battery of the battery swap cabinet is called in combination with the battery health state parameters to determine the target charging mode cluster; the aging evaluation module: the battery of the battery swap cabinet is evaluated according to the target charging mode cluster, and a battery aging evaluation result is generated.

[0007] The one or more technical solutions provided in the present application have at least the following technical effects or advantages: Firstly, the charging period of the battery of the battery swap cabinet is analyzed, and multi-dimensional charging characteristics are extracted for evaluating the health state of the battery. Subsequently, through detailed analysis of the battery charging process, charging mode clusters are constructed, and battery health analysis is performed based on these clusters to obtain battery health state parameters. Then, the charging data of the battery is monitored in real time, and the health state parameters are matched with the plurality of charging mode clusters to determine the charging mode cluster most suitable for the current battery. Finally, the battery is evaluated according to the target charging mode cluster, and a detailed aging evaluation result is generated to help predict the service life of the battery. BRIEF DESCRIPTION OF DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0009] Figure 1 The flowchart of the battery aging evaluation method of the battery swap cabinet based on charging mode clustering provided by the embodiments of the present application.

[0010] Figure 2 The structure diagram of the battery aging evaluation system of the battery swap cabinet based on charging mode clustering provided by the embodiments of the present application.

[0011] Explanation of reference signs: charging cycle analysis module 11, battery health analysis module 12, matching determination module 13, and aging evaluation module 14. DETAILED DESCRIPTION

[0012] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the specific embodiments, structures, features and effects according to the present application will be described in detail below with reference to the drawings and preferred embodiments.

[0013] Embodiment one, as shown in Figure 1 The application provides a battery aging evaluation method for battery swap cabinets based on charging mode clustering. The method comprises the following steps: According to the charging cycle of the battery swap cabinet, the charging cycle analysis of the battery swap cabinet battery is performed, and a multi-dimensional charging feature is constructed.

[0014] In the embodiments of the application, the charging cycle of the battery swap cabinet refers to the complete process of the battery from the beginning of charging to the end of charging. The system first collects the key parameters such as voltage, current and temperature of the battery in the charging process in real time according to the charging cycle of the battery swap cabinet, and records the changes of these parameters over time. Subsequently, the collected key parameters are subjected to outlier rejection and missing value filling to obtain available key parameter sequences. Then, the processed key parameters are subjected to charging cycle analysis to construct a complete charging cycle state parameter. Based on the charging cycle state parameter, multi-stage analysis is performed to form a multi-dimensional charging feature. The multi-dimensional charging feature can comprehensively reflect the state and performance of the battery in each charging cycle, providing accurate data support for battery health analysis, and helping subsequent aging evaluation and life prediction.

[0015] Further, according to the charging cycle of the battery swap cabinet, the charging cycle analysis of the battery swap cabinet battery is performed, and a multi-dimensional charging feature is constructed. The method comprises the following steps: According to the charging cycle of the battery swap cabinet, the charging cycle analysis of the battery swap cabinet battery is performed, and a multi-dimensional charging feature is constructed. The method comprises the following steps:

[0016] Preferably, during the battery charging process of the battery swap cabinet, first, the battery of the battery swap cabinet is analyzed in time sequence according to the charging period, that is, the voltage value at each time point from the beginning to the end of the battery charging process is recorded to form an initial voltage sequence; the current value at each time point during the charging process is recorded to form an initial current sequence; the temperature value at each time point during the charging process is recorded to form an initial temperature sequence. By integrating these initial sequences into a set, a charging time sequence data is formed, which provides basic time sequence data for subsequent charging cycle analysis. Subsequently, due to the interference of external factors or abnormal fluctuations of the equipment during the charging process, the data at some time may suddenly change, which cannot reflect the normal charging state. Therefore, in order to improve the accuracy and reliability of the data, the initial voltage sequence, the initial current sequence and the initial temperature sequence in the charging time sequence data are analyzed for interference based on the normal range of the preset voltage, current and temperature, the data points outside the normal range are identified, and these data points are defined as abnormal mutation points, which represent external interference or sudden problems in the battery charging process. Then, for the detected abnormal mutation points, the corresponding time period data is removed from the initial voltage sequence, the initial current sequence and the initial temperature sequence to generate a data missing sequence, and then a linear interpolation algorithm is used to fill in the missing data part according to the values of the adjacent known data points to generate complete voltage sequence, current sequence and temperature sequence. Then, the filled voltage, current and temperature sequences will be used for further charging cycle analysis, that is, according to the charging period, the same time sequence analysis, interference analysis, data removal, data filling and other processes are performed on each complete charging cycle from the full charge of the battery to the complete discharge of the battery, so as to obtain the voltage sequence, current sequence and temperature sequence of multiple charging cycles. Then, these voltage sequence, current sequence and temperature sequence are integrated and stored to construct a charging cycle state parameter. Finally, based on the charging cycle state parameter, the battery charging process is divided into multiple stages, such as constant current charging stage, constant voltage charging stage and constant temperature charging stage. For each stage, corresponding time domain analysis is performed to extract the time domain features of each stage, thereby constructing multi-dimensional charging features. These charging features not only reflect the health status of the battery, but also reveal the performance of the battery under different charging modes, providing rich data support and accurate analysis results for battery aging evaluation and health analysis.

[0017] Further, based on the charging cycle state parameter, a multi-stage analysis is performed to construct the multi-dimensional charging features, and the method comprises: The charging cycle state parameter is combined with the voltage sequence to perform stage division, a constant voltage charging stage is determined, the battery of the battery swap cabinet is analyzed in time domain according to the constant voltage charging stage, and a first time domain feature is obtained; the charging cycle state parameter is combined with the current sequence to perform stage division, a constant current charging stage is determined, the battery of the battery swap cabinet is analyzed in time domain according to the constant current charging stage, and a second time domain feature is obtained; the charging cycle state parameter is combined with the temperature sequence to perform stage division, a constant temperature charging stage is determined, the battery of the battery swap cabinet is analyzed in time domain according to the constant temperature charging stage, and a third time domain feature is obtained; and the first time domain feature, the second time domain feature and the third time domain feature are associated and integrated to construct the multi-dimensional charging feature.

[0018] Optionally, after obtaining the charging cycle state parameter, the charging cycle state parameter is combined with the corresponding voltage sequence, current sequence and temperature sequence to perform stage division. In this process, when the voltage sequence of the battery reaches a set constant value, it indicates that the constant voltage charging state is entered, that is, the current gradually decreases, and the battery continues to charge at a constant voltage. At this time, the data corresponding to the time in the charging cycle state parameter is intercepted according to the time stamp corresponding to the constant voltage charging state in the voltage sequence, and the constant voltage charging stage is determined. When the current sequence of the battery reaches a set constant value, it indicates that the constant current charging state is entered. At this time, the same method is used to intercept the data corresponding to the time in the charging cycle state parameter, and the constant current charging stage is determined. When the temperature sequence of the battery reaches a set constant value, it indicates that the constant temperature charging state is entered. At this time, the same method is also used to intercept the data corresponding to the time in the charging cycle state parameter, and the constant temperature charging stage is determined. After determining the different stages of charging, for the constant voltage charging stage, the battery maintains a constant voltage in this stage, and the current gradually decreases. Therefore, when the constant voltage charging stage is analyzed in time domain, the changes of the current and the temperature with time are understood, and the current decay rate, the charging time length, the constant voltage charging rate, the temperature change rate and other data are calculated to constitute the first time domain feature. For the constant current charging stage, the battery maintains a constant current in this stage, and the voltage gradually increases with time. Therefore, when the constant current charging stage is analyzed in time domain, the changes of the voltage and the temperature with time are understood, and the voltage growth rate, the charging time length, the constant current charging rate, the temperature change rate and other data are calculated to constitute the second time domain feature. For the constant temperature charging stage, the temperature remains stable and fluctuates within a certain range. At this time, the temperature fluctuation amplitude and the temperature change rate and other data are calculated by analyzing the temperature sequence to constitute the third time domain feature. Finally, the first time domain feature, the second time domain feature and the third time domain feature are associated and integrated to constitute the multi-dimensional charging feature. These multi-dimensional charging features not only reveal the health state of the battery, but also help to evaluate the aging state and charging performance of the battery through comprehensive analysis of different stages.

[0019] Based on the multi-dimensional charging characteristics, clustering learning is performed to construct a plurality of charging mode clusters, battery health analysis is performed on the battery cabinet battery based on the plurality of charging mode clusters, and a battery health state parameter is obtained.

[0020] In one embodiment, the multi-dimensional charging characteristics extracted from the battery during the charging process are taken as input data, clustering learning is performed, and the charging characteristics are divided into a plurality of charging mode clusters, each cluster representing a specific charging mode of the battery during the charging process, and the charging characteristics of the batteries in each cluster are similar, indicating that these batteries operate in similar charging environments. Subsequently, each charging mode cluster is input into a battery health assessment model for analysis, and the battery health assessment model is based on a deep neural network and is trained using historical charging characteristics and historical state parameters. The training steps usually include forward propagation, loss calculation (such as mean square error), back propagation, parameter optimization (such as Adam optimizer), etc. After the battery health assessment model receives the charging mode cluster, it will perform health analysis on the charging characteristics in the charging mode cluster based on the learned knowledge, and generate battery health state parameters for each charging mode. These battery health state parameters usually include remaining capacity, charging efficiency, health state index, internal resistance, maximum charging voltage, and discharge depth, which can reflect the actual health status of the battery and provide data support for subsequent battery aging assessment.

[0021] Further, based on the multi-dimensional charging characteristics, clustering learning is performed to construct a plurality of charging mode clusters, and the method comprises: The multi-dimensional charging characteristics are processed to obtain low-dimensional feature vectors; based on the low-dimensional feature vectors, clustering learning is performed to determine K clustering numbers, K being a positive integer greater than 0; based on the K clustering numbers, clustering division is performed to construct a plurality of initial data clusters; the plurality of initial data clusters are traversed for optimization verification to construct the plurality of charging mode clusters.

[0022] Preferably, for the obtained multi-dimensional charging features, common dimension reduction techniques such as principal component analysis or t-SNE are used to convert the multi-dimensional charging features into low-dimensional feature vectors. Taking principal component analysis as an example, the low-dimensional feature vectors are obtained by eigenvalue decomposition of the covariance matrix, and usually the first k principal component components with the largest variance contribution rate. Through dimension reduction processing, the computational complexity can be reduced, and the efficiency and accuracy of subsequent analysis can be improved. Subsequently, based on these low-dimensional feature vectors, the silhouette coefficient method is used to determine the specified cluster number K, which is a positive integer greater than 0. Then, according to the determined cluster number K, K cluster centers are randomly initialized from the low-dimensional feature vectors, and the Euclidean distance between each low-dimensional feature vector and the K cluster centers is calculated. Then, each low-dimensional feature vector is assigned to the nearest cluster, at this time, the battery charging features in each cluster are similar, representing a state of the charging mode. Next, the mean of all low-dimensional feature vectors in the cluster is calculated as the new cluster center, and the low-dimensional feature vectors are re-assigned to the nearest cluster. This process is repeated until the cluster center no longer changes significantly or reaches the set maximum number of iterations, forming multiple initial data clusters. Then, since multiple initial data clusters may contain noise or unreasonable grouping, data center analysis, intra-cluster consistency check and other steps are used to optimize and verify these clusters to gradually improve the quality of clustering and ensure that each cluster represents a charging mode. The optimized results will constitute multiple charging mode clusters, which can clearly describe the impact of different charging modes on battery health and provide effective basis for subsequent battery health analysis and aging evaluation.

[0023] Further, based on the low-dimensional feature vectors, the K cluster number is determined by clustering learning, and the method comprises: Based on the low-dimensional feature vectors, a fuzzy boundary is divided, and a cluster number candidate threshold is set. According to the cluster number candidate threshold, the low-dimensional feature vectors are calculated to generate a plurality of silhouette coefficients, and the silhouette coefficients have a corresponding relationship with the cluster number. The plurality of silhouette coefficients are arranged in descending order, and the first-order silhouette coefficient is extracted as the K cluster number.

[0024] Optionally, before the cluster analysis, in order to determine the optimal K value, first, the Gaussian membership function is used to calculate the membership degree of each low-dimensional feature vector and all candidate cluster centers, and a membership matrix is constructed to reflect the relationship between the low-dimensional feature vector and the cluster. Subsequently, based on the membership matrix, combined with domain knowledge or the natural distribution of data, the candidate threshold of the number of clusters is set, so that the overlapping area of different clusters is minimized, and the data points in the cluster are tightly clustered. Generally, if a certain candidate K value leads to excessive overlap between clusters or the membership degree of data points is relatively uniform, this K value may not be suitable; when the membership degree is concentrated and the data points in the cluster are tightly clustered, the K value is more appropriate. Then, for each low-dimensional feature vector, according to the determined candidate threshold of the number of clusters, the corresponding silhouette coefficient is calculated. The value of these silhouette coefficients ranges from -1 to 1. The closer the value is to 1, the better the clustering quality, the more tightly the data points in the cluster are clustered, and the more separated the data points between clusters are. When the value is close to 0, it means that the data points may be located at the boundary of the cluster, and when the value is close to -1, it means that the data points may be incorrectly assigned to the wrong cluster. After obtaining the silhouette coefficients of all low-dimensional feature vectors, for each possible K value, the average value of the silhouette coefficients of all low-dimensional feature vectors is calculated to obtain the average value of the silhouette coefficients of each cluster number K. By arranging the average values of the silhouette coefficients under different K values in descending order, the K value corresponding to the maximum average value of the silhouette coefficients is selected as the optimal number of clusters, which provides a reasonable and reliable basis for subsequent cluster learning and data analysis, thereby ensuring the accuracy of the cluster analysis and the quality of the cluster results.

[0025] Further, traversing the plurality of initial data clusters for optimization verification, constructing the plurality of charging mode clusters, the method comprising: Based on the plurality of initial data clusters, data center analysis is performed to determine a plurality of data cluster centers. Data point identification is performed by traversing the plurality of initial data clusters, and a plurality of cluster sample points are extracted according to the identification results. According to the plurality of initial data clusters, the plurality of cluster sample points are matched with the plurality of data cluster centers to generate a plurality of data point combinations, each data point combination in the plurality of data point combinations having and only having one data cluster center and a plurality of cluster sample points. Based on the plurality of data cluster centers and the plurality of cluster sample points, verification is performed, and according to the verification results, the plurality of initial data clusters are optimized to construct the plurality of charging mode clusters.

[0026] Optionally, after obtaining the plurality of initial data clusters, data center analysis is performed on each initial data cluster to calculate the mean value of each initial data cluster and determine the plurality of data cluster centers. Subsequently, the plurality of initial data clusters are traversed, the distance between each data point traversed and the data cluster center is calculated, and these distances are used to label these data points, and then the points closer to the cluster center are selected as cluster sample points according to the labels. Then, the cluster sample points are matched with the data cluster center to generate a plurality of data point combinations, each data point combination containing a cluster center and a plurality of cluster sample points, for further verification and optimization. Then, the data clusters are verified and optimized according to the data cluster center and the cluster sample points in each data point combination, by verifying whether the similarity between the cluster sample points and the cluster center in each data point combination is high enough to check whether the initial clustering is effective. Finally, according to the verification result, the boundaries of the clusters are adjusted, and the data points that do not meet the requirements are re-assigned to other clusters to avoid the overlap of clusters or the wrong division of sample points, thereby obtaining the final plurality of charging mode clusters, which provides an accurate basis for subsequent battery aging assessment, life prediction and other analyses.

[0027] Further, based on the plurality of data cluster centers and the plurality of cluster sample points, verification is performed, and according to the verification result, the plurality of initial data clusters are optimized to construct the plurality of charging mode clusters, the method comprising: Based on the plurality of data point combinations, the average distance between the plurality of data cluster centers and the plurality of cluster sample points is calculated to generate a plurality of average distance values, and the total average distance is calculated. When the average distance value in any data point combination exceeds Q times of the total average distance, the data point combination is identified as a loose cluster, and Q is a positive integer greater than 1. When the loose cluster exists, the plurality of data point combinations are clustered and optimized, the plurality of initial data clusters are updated, and the plurality of charging mode clusters are constructed.

[0028] Optionally, for each data point combination, the Euclidean distance between the data cluster center and all cluster class sample points is calculated, and the average distance value is obtained, which is used to measure the tightness of the cluster. The smaller the value, the more concentrated the sample points in the cluster, and the higher the quality of clustering. Then, a total average distance is calculated for all data point combinations, which is the average of the average distances between all cluster centers and sample points, representing the overall tightness of all clusters. Then, a threshold Q is defined to identify loose clusters. When any average distance value exceeds Q times the total average distance, the cluster is considered a loose cluster and is identified. For the initial data clusters identified as loose clusters, some data points in the loose cluster are reassigned to a more similar cluster, or the cluster boundary is reevaluated, and the cluster center position is updated by recalculating the mean of the data points in the cluster to ensure that the cluster center represents the center position of the data points in the cluster. After the above optimization and update steps, the final multiple charging mode clusters will be more compact and have better discrimination. These charging mode clusters can more accurately reflect the state of health of the battery under different charging modes, providing more reliable data support for subsequent battery aging assessment, performance analysis, etc.

[0029] The real-time charging data of the battery of the battery swap cabinet is retrieved and matched with the battery state of health parameters to determine the target charging mode cluster.

[0030] In one embodiment, the system monitors and obtains various real-time charging data of the battery of the battery swap cabinet in the current charging process in real time. These real-time charging data typically include information such as the voltage, current, temperature, remaining capacity, charging efficiency, state of health index, internal resistance of the battery, etc., which can reflect the state of the battery in a specific charging process. Then, the real-time charging data are matched with the battery state of health parameters, the similarity between the real-time charging data and the battery state of health parameters is calculated by Euclidean distance, and the nearest set of battery state of health parameters is selected. Then, the charging mode cluster corresponding to this set of battery state of health parameters is obtained, and this charging mode cluster is used as the target charging mode cluster. This target charging mode cluster best reflects the state of health and charging characteristics of the current battery in the charging process, ensuring the accuracy of subsequent battery aging assessment.

[0031] The battery of the battery swap cabinet is subjected to battery aging assessment according to the target charging mode cluster, and a battery aging assessment result is generated.

[0032] In one embodiment, the battery health state data and charging characteristics in the target charging mode cluster are first used to evaluate the battery in the battery swap cabinet for multiple charging cycles, and the evaluation values are then organized in chronological order into a health state history sequence, reflecting the changing trend of the battery in different charging cycles. Subsequently, the battery health state parameters of the target charging mode cluster are trained by a regression algorithm to generate a baseline aging curve, which represents the normal aging trend of the battery. Then, the health state history sequence is compared with the baseline aging curve, and the time axis is adjusted using a dynamic time warping algorithm to accurately align the battery health data with the baseline aging curve. Then, the remaining useful life of the battery, i.e., the time the battery can continue to be effectively used under the current state, is predicted based on the comparison result, thereby generating a complete battery aging evaluation result including the battery health state, aging rate and remaining useful life, etc., providing a decision basis for subsequent battery replacement, maintenance and management.

[0033] Further, the battery aging evaluation of the battery swap cabinet battery according to the target charging mode cluster is performed to generate a battery aging evaluation result, and the method comprises: Based on the target charging mode cluster, the battery in the battery swap cabinet is evaluated for multiple charging cycles to obtain health state estimation values for the multiple charging cycles; the health state estimation values for the multiple charging cycles are sequentially processed according to the charging time sequence to construct a health state history sequence; data regression training is performed according to the target charging mode cluster to set a baseline aging curve; the health state history sequence is compared with the baseline aging curve, and dynamic time warping is performed according to the comparison result to calculate battery aging rate data; attenuation evaluation is performed according to the battery aging rate data to construct a battery aging trajectory; the remaining useful life prediction value of the battery is obtained based on the battery aging trajectory for life prediction of the battery in the battery swap cabinet, and the battery remaining useful life prediction value is added to the battery aging evaluation result.

[0034] Preferably, based on the determined target charging mode cluster, the health state evaluation model corresponding to the target charging mode is called, which is trained based on the historical data of the charging mode, and a deep neural network can also be used as the model framework, and the training steps are the same as described above. After calling the corresponding health state evaluation model, the key parameters of the battery swap cabinet battery in multiple charging cycles, such as voltage, current, temperature, etc., are input into this model, and a health state estimate value is calculated comprehensively to reflect the performance of the battery in multiple charging cycles. Subsequently, the health state estimate values of all charging cycles are arranged in chronological order to form a health state history sequence to reflect the health changes of the battery at different time points and charging cycles. After constructing the health state history sequence, the system uses the historical data corresponding to the target charging mode cluster for data regression training. In this process, taking the exponential regression model as an example, the time and health state values in the historical data are input into the exponential regression model, the least squares method is used to minimize the loss function, and the model parameters are updated to fit the data until the loss function reaches the minimum value. After completing the regression training, the baseline aging curve of the battery is drawn according to the parameters in the exponential regression model, which is a theoretical standard curve representing the aging process of the battery under typical use conditions. Then, the dynamic time warping algorithm is used to align the health state history sequence with the generated baseline aging curve, determine the difference between them, and calculate the aging rate of the battery, i.e. the speed of battery degradation, to help predict the future performance and aging process of the battery under the same charging mode. Then, based on the calculated battery aging rate data, the health state history sequence is traversed from the starting point for decay evaluation, and the battery aging trajectory is drawn, which depicts how the health state and performance of the battery gradually decline over time, and can be used to predict the future degradation process of the battery. Finally, according to the battery aging trajectory, the position corresponding to the current health state estimate value is located to understand which stage the battery is currently in, such as the early degradation stage, the middle degradation stage, or the near failure stage, and then the degradation model belonging to the current stage is used to estimate the remaining useful life of the battery according to the current health state. These degradation models are all built based on long short-term memory networks. After obtaining the battery remaining useful life prediction value, the battery remaining useful life prediction value is added to the battery aging evaluation result to provide a comprehensive evaluation for the future use of the battery, ensuring the efficient and reliable operation of the battery swap cabinet battery.

[0035] Further, the decay evaluation according to the battery aging rate data to construct the battery aging trajectory comprises: Take the last data point in the health state history sequence as the starting point, and take the battery aging rate data as the attenuation coefficient; based on the attenuation coefficient, the increase in the charging cycle is calculated by traversing the health state history sequence from the starting point, and the battery attenuation decline data is obtained; the battery attenuation decline data is visualized to construct the battery aging trajectory.

[0036] Optionally, in the battery aging evaluation process, first, the last data point in the health state history sequence is selected as the starting point, which represents the health status of the battery at the current time, and then the battery aging rate data is taken as the attenuation coefficient to simulate the change of the health status of the battery over time and predict the future decline trend. Subsequently, the health state history sequence is traversed in reverse, and each data point is processed one by one. At each data point, the number of charging cycles is increased, and the health state of the battery is calculated based on the attenuation coefficient. Specifically, for each charging cycle, the health state of the battery will decrease by a certain percentage according to the attenuation coefficient, and the attenuation calculation formula is the current health state estimate minus the product of the attenuation coefficient and the health state estimate. During the reverse traversal process, the health state of the battery will gradually decline, generating attenuation decline data after each charging period. This process will continue until the starting point of the health state history sequence is traversed, thereby obtaining a complete battery attenuation decline data set that records the health status change trend of the battery in multiple charging periods, providing detailed basis for subsequent battery aging trajectory construction and life prediction. Subsequently, the battery attenuation decline data set will be visualized and plotted as an attenuation curve as the final battery aging trajectory, which shows the decline process of the battery from the current state to the end-of-life point, helping to predict the remaining useful life of the battery.

[0037] In summary, the embodiments of the present application have at least the following technical effects: First, the battery of the battery swap cabinet is analyzed for charging cycle according to the charging period of the battery swap cabinet, and multi-dimensional charging features are constructed. Then, based on the multi-dimensional charging features, a plurality of charging mode clusters are constructed, and battery health analysis is performed on the battery of the battery swap cabinet based on the plurality of charging mode clusters to obtain battery health state parameters. Then, the real-time charging data of the battery of the battery swap cabinet is retrieved in combination with the battery health state parameters to match and determine the target charging mode cluster. Finally, the battery of the battery swap cabinet is evaluated for battery aging according to the target charging mode cluster to generate a battery aging evaluation result. The technical problem of low evaluation accuracy and poor real-time performance caused by the diversity of charging modes in the existing battery aging evaluation method is solved, and the technical effects of realizing fine battery health state analysis and dynamic matching based on charging mode clustering, improving the accuracy, scene adaptability and real-time responsiveness of the battery aging evaluation result are achieved.

[0038] Embodiment two, based on the same inventive concept as the battery aging evaluation method based on charging mode clustering in the preceding embodiment, as shown in Figure 2 The present application provides a battery aging evaluation system based on charging mode clustering, which comprises: The charging cycle analysis module 11 performs charging cycle analysis on the battery of the battery swap cabinet according to the charging period of the battery swap cabinet, and constructs multi-dimensional charging features; the battery health analysis module 12 performs clustering learning based on the multi-dimensional charging features, constructs multiple charging mode clusters, performs battery health analysis on the battery of the battery swap cabinet based on the multiple charging mode clusters, and obtains battery health status parameters; the matching determination module 13 retrieves real-time charging data of the battery of the battery swap cabinet in combination with the battery health status parameters to determine the target charging mode cluster; and the aging evaluation module 14 performs battery aging evaluation on the battery of the battery swap cabinet according to the target charging mode cluster, and generates a battery aging evaluation result.

[0039] Further, the charging cycle analysis module 11 is used to perform the following method: The charging cycle analysis module 11 is used to perform the following method: Further, the charging cycle analysis module 11 is used to perform the following method:

[0040] Further, the charging cycle analysis module 11 is used to perform the following method: Further, the charging cycle analysis module 11 is used to perform the following method:

[0041] Further, the battery health analysis module 12 is configured to perform the following method: dimension reduction processing is performed on the multi-dimensional charging features to obtain a low-dimensional feature vector; clustering learning is performed based on the low-dimensional feature vector to determine a number of K clusters, K being a positive integer greater than 0; clustering division is performed based on the number of K clusters to construct a plurality of initial data clusters; and optimization verification is performed by traversing the plurality of initial data clusters to construct the plurality of charging mode clusters.

[0042] Further, the battery health analysis module 12 is configured to perform the following method: fuzzy boundary division is performed based on the low-dimensional feature vector to set a cluster number candidate threshold; contour calculation is performed on the low-dimensional feature vector according to the cluster number candidate threshold to generate a plurality of contour coefficients, the contour coefficients having a corresponding relationship with the cluster number; and the plurality of contour coefficients are arranged in descending order, and the contour coefficient at the first order is extracted as the number of K clusters.

[0043] Further, the battery health analysis module 12 is configured to perform the following method: data center analysis is performed based on the plurality of initial data clusters to determine a plurality of data cluster centers; data point identification is performed by traversing the plurality of initial data clusters, and a plurality of cluster sample points are extracted according to the identification result; the plurality of cluster sample points are matched with the plurality of data cluster centers according to the plurality of initial data clusters to generate a plurality of data point combinations, each data point combination in the plurality of data point combinations having and only having one data cluster center and a plurality of cluster sample points; verification is performed based on the plurality of data cluster centers and the plurality of cluster sample points, and the plurality of initial data clusters are optimized according to the verification result to construct the plurality of charging mode clusters.

[0044] Further, the battery health analysis module 12 is configured to perform the following method: Based on the plurality of data point combinations, the average distance between the plurality of data cluster centers and the plurality of cluster sample points is calculated to generate a plurality of average distance values; the total average distance is calculated, and when the average distance value in any data point combination exceeds Q times of the total average distance, the data point combination is identified as a loose cluster, Q being a positive integer greater than 1; when the loose cluster exists, clustering optimization is performed on the plurality of data point combinations, the plurality of initial data clusters are updated, and the plurality of charging mode clusters are constructed.

[0045] Further, the aging evaluation module 14 is configured to perform the following method: The battery aging evaluation module 14 is configured to perform the following method:

[0046] Further, the aging evaluation module 14 is configured to perform the following method: The last data point in the health state historical sequence is taken as a starting point, and the battery aging rate data is taken as an attenuation coefficient; the health state historical sequence is traversed from the starting point based on the attenuation coefficient to perform charging cycle increase calculation, and battery attenuation drop data is obtained; the battery attenuation drop data is visualized to construct the battery aging trajectory.

[0047] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above with a preferred embodiment, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content to obtain equivalent embodiments with equivalent changes, without departing from the technical solution of the present application. Any modification, equivalent change and modification of the above embodiments based on the technical essence of the present application are still within the scope of the technical solution of the present application.

Claims

1. A battery aging assessment method for battery swapping cabinets based on charging mode clustering, characterized in that, The method includes: Based on the charging cycle of the battery swapping cabinet, a charging cycle analysis of the battery swapping cabinet is performed to construct multi-dimensional charging characteristics; Based on the multidimensional charging features, cluster learning is performed to construct multiple charging mode clusters. Based on the multiple charging mode clusters, battery health analysis is performed on the battery swapping cabinet battery to obtain battery health status parameters. The real-time charging data of the battery in the battery swapping cabinet is retrieved and combined with the battery health status parameters to match and determine the target charging mode cluster. The battery aging assessment of the battery swapping cabinet is performed according to the target charging mode cluster, and the battery aging assessment results are generated.

2. The battery aging assessment method for battery swapping cabinets based on charging pattern clustering as described in claim 1, characterized in that, The charging cycle analysis of the batteries in the battery swapping cabinet is performed according to the charging cycle of the cabinet to construct multi-dimensional charging characteristics. The methods include: The battery in the battery swapping cabinet is analyzed according to the charging cycle of the battery swapping cabinet to extract charging time sequence data, which includes initial voltage sequence, initial current sequence and initial temperature sequence. Interference analysis is performed by traversing the initial voltage sequence, the initial current sequence, and the initial temperature sequence to identify abnormal abrupt changes; Based on the anomalous mutation points, data is removed from the initial voltage sequence, the initial current sequence, and the initial temperature sequence to generate a data missing sequence; The missing data sequence is filled in using a linear interpolation algorithm to generate voltage, current, and temperature sequences. The voltage sequence, current sequence, and temperature sequence are subjected to charging cycle analysis to construct charging cycle state parameters; Based on the charging cycle state parameters, a multi-stage analysis is performed to construct the multi-dimensional charging characteristics.

3. The battery aging assessment method for battery swapping cabinets based on charging mode clustering as described in claim 2, characterized in that, The method for constructing the multidimensional charging characteristics based on the charging cycle state parameters through multi-stage analysis includes: Based on the charging cycle state parameters and the voltage sequence, the constant voltage charging stage is determined by dividing the stage into phases. The battery in the battery swapping cabinet is analyzed in the time domain according to the constant voltage charging stage to obtain the first time domain feature. Based on the charging cycle state parameters and the current sequence, the constant current charging stage is determined by dividing the process into stages. Perform time-domain analysis on the battery swapping cabinet battery according to the constant current charging stage to obtain the second time-domain characteristics; Based on the charging cycle state parameters and the temperature sequence, the constant temperature charging stage is determined by dividing the stage into phases. A time-domain analysis of the battery in the battery swapping cabinet is performed according to the constant temperature charging stage to obtain the third time-domain feature. The first time-domain feature, the second time-domain feature, and the third time-domain feature are correlated and integrated to construct the multidimensional charging feature.

4. The battery aging assessment method for battery swapping cabinets based on charging mode clustering as described in claim 1, characterized in that, Clustering learning is performed based on the aforementioned multidimensional charging features to construct multiple charging pattern clusters. The method includes: The multidimensional charging features are subjected to dimensionality reduction processing to obtain a low-dimensional feature vector; Clustering learning is performed based on the low-dimensional feature vectors to determine the number of K clusters, where K is a positive integer greater than 0; Based on the K clustering numbers, clustering is performed to construct multiple initial data clusters; The multiple initial data clusters are traversed for optimization and verification to construct the multiple charging mode clusters.

5. The battery aging assessment method for battery swapping cabinets based on charging mode clustering as described in claim 4, characterized in that, Clustering learning is performed based on the low-dimensional feature vectors to determine the number of K clusters. The method includes: Based on the low-dimensional feature vector, fuzzy boundary division is performed, and a candidate threshold for the number of clusters is set. The low-dimensional feature vector is contoured according to the candidate threshold for the number of clusters to generate multiple contour coefficients, and the contour coefficients are related to the number of clusters. The multiple silhouette coefficients are sorted in descending order, and the first-order silhouette coefficient is extracted as the number of K clusters.

6. The battery aging assessment method for battery swapping cabinets based on charging pattern clustering as described in claim 4, characterized in that, The method involves iterating through the multiple initial data clusters for optimization and verification, and constructing the multiple charging mode clusters, including: Data center analysis is performed based on the multiple initial data clusters to determine multiple data cluster centers; The multiple initial data clusters are traversed to identify data points, and multiple cluster sample points are extracted based on the identification results. According to the multiple initial data clusters, the multiple cluster sample points are matched with the multiple data cluster centers to generate multiple data point combinations. Each data point combination has one and only one data cluster center and multiple cluster sample points. Based on the verification of the multiple data cluster centers and the multiple cluster sample points, the multiple initial data clusters are optimized according to the verification results to construct the multiple charging mode clusters.

7. The battery aging assessment method for battery swapping cabinets based on charging pattern clustering as described in claim 6, characterized in that, The method involves verifying the multiple data cluster centers and the multiple cluster sample points, optimizing the multiple initial data clusters based on the verification results, and constructing the multiple charging mode clusters. Based on the combination of the multiple data points, the average distance between the centers of the multiple data clusters and the sample points of the multiple clusters is calculated, and multiple average distance values ​​are generated. Calculate the total average distance. If the average distance within any data point combination exceeds Q times the total average distance, then the data point combination is identified as a loose cluster, where Q is a positive integer greater than 1. When the loose clusters exist, cluster optimization is performed on the combination of the multiple data points, the multiple initial data clusters are updated, and the multiple charging mode clusters are constructed.

8. The battery aging assessment method for battery swapping cabinets based on charging pattern clustering as described in claim 1, characterized in that, The battery aging assessment of the battery swapping cabinet battery is performed according to the target charging mode cluster, and the battery aging assessment results are generated. The method includes: Based on the target charging mode cluster, the battery swapping cabinet battery is evaluated for charging in multiple charging cycles to obtain health status estimates for multiple charging cycles. The health status estimates of the multiple charging cycles are serialized according to the charging time sequence to construct a health status history sequence. Data regression training is performed according to the target charging mode cluster, and a baseline aging curve is set. The health status history sequence is compared with the baseline aging curve, and dynamic time normalization is performed based on the comparison results to calculate the battery aging rate data. Based on the battery aging rate data, a degradation assessment is performed to construct a battery aging trajectory. Based on the battery aging trajectory, the battery life of the battery in the battery swapping cabinet is predicted to obtain the predicted value of the remaining useful life of the battery, and the predicted value of the remaining useful life of the battery is added to the battery aging assessment result.

9. The battery aging assessment method for battery swapping cabinets based on charging pattern clustering as described in claim 8, characterized in that, Based on the battery aging rate data, a degradation assessment is performed to construct a battery aging trajectory. The method includes: The last data point in the health status history sequence is taken as the starting point, and the battery aging rate data is taken as the decay coefficient. Based on the attenuation coefficient, the charging cycle is increased by traversing the historical sequence of the health state from the starting point to obtain battery attenuation data. The battery degradation data is visualized to construct the battery aging trajectory.

10. A battery aging assessment system for battery swapping cabinets based on charging mode clustering, characterized in that, The system is used to implement the battery aging assessment method for battery swapping cabinets based on charging pattern clustering as described in any one of claims 1-9, the system comprising: Charging cycle analysis module: Performs charging cycle analysis on the batteries in the battery swapping cabinet according to the charging cycle of the battery swapping cabinet, and constructs multi-dimensional charging characteristics; Battery health analysis module: Based on the multi-dimensional charging features, cluster learning is performed to construct multiple charging mode clusters. Based on the multiple charging mode clusters, battery health analysis is performed on the battery in the battery swapping cabinet to obtain battery health status parameters. Matching and Determination Module: Retrieves real-time charging data of the battery in the battery swapping cabinet and combines it with the battery health status parameters to match and determine the target charging mode cluster. Aging assessment module: Performs battery aging assessment on the battery in the battery swapping cabinet according to the target charging mode cluster, and generates battery aging assessment results.

Images