Battery screening method, device, equipment and storage medium based on dynamic and static characteristics

Abstract

The present invention belongs to the field of battery technology, and discloses a battery screening method, device, equipment and storage medium based on dynamic and static features. The method includes: constructing an incremental capacity curve based on the sample parameters of the retired battery, and extracting the peak characteristics of the curve based on the incremental capacity curve; performing battery static performance consistency screening on the retired battery based on the peak characteristics of the curve and the corresponding battery internal resistance to obtain the screened target retired battery; performing dynamic feature extraction based on the operating data of the target retired battery, determining the dynamic performance characteristics of the battery and implementing feature characterization in a graphical manner; inputting the battery dynamic performance characteristic image and the battery operating status characteristic image into the image classification model to complete battery screening based on dynamic and static features. The above method realizes the consistency sorting of retired batteries, thereby providing a reference for battery recycling and improving the consistency of retired battery sorting.

Description

Technical Field

[0001] The present invention relates to the field of battery technology, and in particular to a battery screening method, device, equipment and storage medium based on dynamic and static characteristics. Background Art

[0002] Lithium batteries, as an efficient and lightweight energy storage solution, are widely used in portable electronic devices, electric vehicles, grid energy storage, and other applications due to their high energy density, long lifespan, and low self-discharge rate. However, batteries typically have a service life of 5 to 8 years. With the continued growth of the electric vehicle market, the number of batteries reaching end-of-life due to performance degradation is increasing. Therefore, effectively disposing of retired batteries has become a pressing challenge. While these batteries may no longer meet the performance requirements of electric vehicles, their average remaining capacity remains as high as 80%, offering significant potential for secondary use. Among existing battery recycling solutions, cascaded reuse is a promising approach for reusing retired batteries. Cascaded reuse involves screening, sorting, and reassembling retired batteries based on performance, allowing them to be reused in applications such as energy storage systems, power tools, and low-speed electric vehicles, thereby extending battery life and improving economic and environmental benefits. However, due to the uneven aging characteristics of retired batteries, performance inconsistencies between individual cells are significant, leading to a pronounced "bucket effect" in battery modules, with inconsistent performance across individual cells, resulting in reduced overall module efficiency. Furthermore, poor consistency can lead to overcharging or over-discharging of some individual cells, causing rapid degradation and creating a vicious cycle. Therefore, it is necessary to develop consistent sorting technology to ensure the economy and safety of the cascade utilization of retired batteries.

[0003] Existing methods primarily rely on parameters such as capacity and internal resistance as performance evaluation metrics, employing clustering methods to achieve battery sorting. However, due to the capacity regeneration effect, the measured capacity of lithium-ion batteries often does not accurately reflect their true performance state. Internal resistance also fluctuates due to factors such as temperature changes. Therefore, relying solely on capacity and internal resistance to assess the consistency of retired batteries is insufficient. To address this, some studies have extracted additional features from battery operating data to characterize performance, including charge and discharge times, voltage, and current averages. However, these features primarily represent the static characteristics of the battery. During battery reassembly, it is crucial not only to ensure the similarity of static metrics but also to maximize the consistency of battery performance during actual operation. Therefore, developing a method that considers dynamic performance is crucial. However, existing methods fail to fully consider these issues, often overlooking the need to screen out inconsistent batteries before classification. Furthermore, traditional feature extraction methods may not fully capture the comprehensive characteristics of battery dynamics. Furthermore, the model's feature extraction capabilities need to be further enhanced.

[0004] The above content is only used to assist in understanding the technical solution of the present invention and does not constitute an admission that the above content is prior art. Summary of the Invention

[0005] The main purpose of the present invention is to provide a battery screening method, device, equipment and storage medium based on dynamic and static characteristics, aiming to solve the technical problem that the performance consistency of retired batteries in the existing technology is not enough to affect the classification effect.

[0006] To achieve the above objectives, the present invention provides a battery screening method based on dynamic and static characteristics, the method comprising the following steps:

[0007] Obtain sample parameters of retired batteries;

[0008] constructing an incremental capacity curve according to the sample parameters of the retired battery, and extracting a curve peak feature based on the incremental capacity curve;

[0009] Performing battery performance consistency screening on the retired batteries according to the peak characteristics of the curve and the corresponding battery internal resistance to obtain screened target retired batteries;

[0010] Extract features based on sample parameters of the target retired battery to determine a battery dynamic performance feature image and a battery operating status feature image;

[0011] The battery dynamic performance characteristic image and the battery operating status characteristic image are input into an image classification model to complete battery screening based on dynamic and static characteristics.

[0012] Optionally, constructing an incremental capacity curve according to sample parameters of the retired battery includes:

[0013] Performing incremental capacity analysis based on the charging process capacity and voltage in the sample parameters to obtain initial incremental capacity data;

[0014] The initial incremental capacity data is filtered by a filtering algorithm to construct an incremental capacity curve;

[0015] The filtering algorithm includes:

[0016] ;

[0017] in is the output after filtering, is the first data points, are the coefficients obtained by polynomial fitting, is the window size.

[0018] Optionally, the performing battery performance consistency screening on the retired batteries according to the curve peak characteristics and the corresponding battery internal resistance includes:

[0019] Generate a screening parameter set according to the peak coordinates in the peak characteristics of the curve and the internal resistance of the battery;

[0020] Performing cluster analysis on the screening parameter set based on a preset domain, a preset neighborhood radius, and a preset minimum density to determine core points, edge points, or noise points;

[0021] The preset fields are: ,in, is the sample point and The Euclidean distance between them, Eps is the preset neighborhood radius;

[0022] The retired batteries are screened for battery performance consistency based on the core points, edge points or noise points.

[0023] Optionally, extracting features based on sample parameters of the target retired battery to determine a battery dynamic performance characteristic image includes:

[0024] Determining a voltage-capacity variation curve according to sample parameters of the target retired battery;

[0025] Normalizing the voltage-capacity change curve to obtain a standardized voltage-capacity change curve;

[0026] Interpolation processing is performed on the standardized voltage-capacity change curve to generate a battery operation status characteristic image of a preset size.

[0027] Optionally, extracting features based on sample parameters of the target retired battery to determine a battery dynamic performance characteristic image includes:

[0028] Performing incremental capacity analysis on the sample parameters of the target retired battery to extract an incremental capacity curve of the target retired battery;

[0029] Performing polar coordinate conversion on the incremental capacity curve to obtain a target polar coordinate sequence;

[0030] Calculating the inner product of the vectors according to the target polar coordinate sequence to obtain a Gram matrix;

[0031] Calculating the Gram angle sum field and the Gram angle difference field based on the target polar coordinate sequence, the calculation scheme includes:

[0032] ;

[0033] ;

[0034] in, is a unit vector, is the target polar coordinate sequence, φi and φj are the corresponding angles in polar coordinates; represents the matrix after Gram angle and field transformation, represents the matrix after Gram angle difference field transformation;

[0035] A battery dynamic performance characteristic image is determined according to the Gram matrix, the Gram angle sum field, and the Gram angle difference field.

[0036] Optionally, inputting the battery dynamic performance characteristic image and the battery operating status characteristic image into an image classification model to complete battery screening based on dynamic and static characteristics includes:

[0037] The battery dynamic performance characteristic image and the battery operating status characteristic image backbone network are used to determine the high-dimensional vector features of the target retired battery, wherein the backbone network includes a dense connection network and a CBAM feature extraction and optimization module;

[0038] Inputting the high-dimensional vector features into a classification layer to determine the category probability;

[0039] Battery screening based on dynamic and static features is completed according to the category probability.

[0040] Optionally, the network structure of the densely connected network is characterized as follows:

[0041] ;

[0042] in, Is a comprehensive function that includes multiple operations, Indicates the 0th to The output, For the The output of the layer;

[0043] The densely connected network includes: ReLu activation function and 3×3 convolution;

[0044] The formulas for the channel attention mechanism and spatial attention mechanism of the CBAM feature extraction and optimization module are:

[0045] ;

[0046] in, represents global average pooling, represents the maximum pooling operation, represents a multilayer perceptron, represents the Sigmoid activation function, Represents a convolution operation.

[0047] In addition, to achieve the above-mentioned purpose, the present invention further proposes a battery screening device based on dynamic and static characteristics, the battery screening device based on dynamic and static characteristics comprising:

[0048] An acquisition module, used to obtain sample parameters of retired batteries;

[0049] a processing module, configured to construct an incremental capacity curve according to sample parameters of the retired battery, and extract a curve peak feature based on the incremental capacity curve;

[0050] The processing module is configured to perform battery performance consistency screening on the retired batteries according to the peak characteristics of the curve and the corresponding battery internal resistance, to obtain screened target retired batteries;

[0051] The processing module is configured to extract features based on sample parameters of the target retired battery to determine a battery dynamic performance feature image and a battery operating status feature image;

[0052] The processing module is used to input the battery dynamic performance characteristic image and the battery operation status characteristic image into the image classification model to complete the battery screening based on dynamic and static characteristics.

[0053] In addition, to achieve the above-mentioned purpose, the present invention also proposes a battery screening device based on dynamic and static features, and the battery screening device based on dynamic and static features includes: a memory, a processor, and a battery screening program based on dynamic and static features stored on the memory and runnable on the processor, and the battery screening program based on dynamic and static features is configured to implement the steps of the battery screening method based on dynamic and static features as described above.

[0054] In addition, to achieve the above-mentioned purpose, the present invention also proposes a storage medium, on which a battery screening program based on dynamic and static features is stored. When the battery screening program based on dynamic and static features is executed by a processor, the steps of the battery screening method based on dynamic and static features as described above are implemented.

[0055] The present invention obtains sample parameters of retired batteries; constructs an incremental capacity curve based on the sample parameters of the retired batteries, and extracts the curve peak features based on the incremental capacity curve; performs battery performance consistency screening on the retired batteries based on the curve peak features and the corresponding battery internal resistance to obtain the screened target retired batteries; performs feature extraction based on the sample parameters of the target retired batteries to determine the battery dynamic performance feature image and the battery operating status feature image; inputs the battery dynamic performance feature image and the battery operating status feature image into an image classification model to complete battery screening based on dynamic and static features. The above method realizes the consistency screening of retired batteries, thereby providing a good data foundation for battery classification and improving the accuracy of retired battery screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0057] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0058] Figure 1 Schematic diagram of the flow of the first embodiment of the battery screening method based on dynamic and static characteristics of the present invention;

[0059] Figure 2 Schematic diagram of the flow of the second embodiment of the battery screening method based on dynamic and static characteristics of the present invention;

[0060] Figure 3 This is a structural block diagram of a first embodiment of a battery screening device based on dynamic and static characteristics of the present invention;

[0061] Figure 4 It is a structural diagram of a battery screening device based on dynamic and static characteristics in a hardware operating environment involved in an embodiment of the present invention.

[0062] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0063] It should be understood that the specific embodiments described herein are merely used to explain the technical solutions of the present application and are not intended to limit the present application.

[0064] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0065] The main solution of the embodiment of the present application is: obtaining sample parameters of retired batteries through an intelligent terminal, such as a server or computer; constructing an incremental capacity curve based on the sample parameters of the retired batteries, and extracting curve peak features based on the incremental capacity curve; performing battery performance consistency screening on the retired batteries based on the curve peak features and the corresponding battery internal resistance to obtain screened target retired batteries; performing feature extraction based on the sample parameters of the target retired batteries to determine a battery dynamic performance feature image and a battery operating status feature image; inputting the battery dynamic performance feature image and the battery operating status feature image into an image classification model to complete battery screening based on dynamic and static features.

[0066] In this embodiment, for ease of description, the following description is made with identification of a smart terminal as the execution subject.

[0067] Existing technologies primarily rely on parameters such as capacity and internal resistance as performance evaluation metrics, employing clustering methods to achieve battery sorting. However, due to the capacity regeneration effect, the measured capacity of lithium-ion batteries often does not accurately reflect their true performance state. Internal resistance also fluctuates due to factors such as temperature changes. Therefore, relying solely on capacity and internal resistance to assess the consistency of retired batteries is insufficient. To address this, some studies have extracted additional features from battery operating data to characterize performance, including charge and discharge times, voltage, and current averages. However, these features primarily represent the static characteristics of the battery. During battery reassembly, it is crucial not only to ensure the similarity of static metrics but also to maximize the consistency of battery performance during actual operation. Therefore, developing a method that considers dynamic performance is crucial. However, existing methods fail to fully consider these issues, often overlooking the need to screen out inconsistent batteries before classification. Traditional feature extraction methods may not fully capture the comprehensive characteristics of battery dynamics. Furthermore, the model's feature extraction capabilities need to be further enhanced.

[0068] This application provides a method for classifying retired batteries. First, this solution innovatively develops a grading framework that integrates static parameter-based screening and dynamic data classification. Using static parameters such as voltage and capacity as input, it uses the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm to achieve efficient initial screening of abnormal batteries. Combined with the designed dynamic performance classification method, it improves the overall consistency of the sorted batteries. Second, this solution innovatively proposes a method for deep mining of dynamic features of retired batteries based on data imaging. By graying the VQ curve, it fully preserves the dynamic characteristics of the battery. By encoding the IC curve into a two-dimensional image through GAF, it captures the battery capacity attenuation trend and peak evolution law, and realizes a graphical representation of the battery aging mechanism. Finally, this solution innovatively constructs a DenseNet-CBAM hybrid neural network model. By combining a densely connected structure with a spatial-channel attention mechanism, it enhances feature reuse efficiency and dynamically optimizes feature weights, ultimately achieving consistent and accurate sorting of retired batteries.

[0069] The two-stage sorting method for retired batteries proposed in this solution, which combines dynamic and static features, effectively screens out batteries with inconsistent performance, and accurately characterizes the dynamic performance of retired batteries through data imaging, ultimately achieving accurate retired battery sorting results. The method proposed in this solution can effectively consider both the static and dynamic performance of batteries, effectively implement battery screening and classification, and ensure the consistency of dynamic and static performance of subsequent battery recycling. Compared with traditional feature extraction-based classification methods, it improves F1-score by 10%, achieving a significant performance improvement. The technology of this solution can be applied to the sorting link of the retired battery recycling process to guide the classification and secondary utilization of retired batteries.

[0070] It can be seen from the above embodiments that the present application realizes the consistency screening of retired batteries, thereby providing a good data foundation for battery classification and improving the accuracy of retired battery screening.

[0071] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, mobile phone, etc., or an electronic device or smart terminal capable of implementing the above functions. The following uses a smart terminal as an example to illustrate this embodiment and the following embodiments.

[0072] Based on this, the embodiment of the present application provides a battery screening method based on dynamic and static characteristics, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the battery screening method based on dynamic and static characteristics of this application.

[0073] In this embodiment, the battery screening method based on dynamic and static characteristics includes the following steps:

[0074] Step S10: Obtain sample parameters of retired batteries.

[0075] It should be noted that retired batteries can be power batteries retired from new energy vehicles (pure electric or hybrid vehicles), including batteries replaced due to capacity degradation, safety risks, or vehicle scrapping. According to the standard, their remaining capacity is usually lower than a certain value of the initial capacity and cannot meet the vehicle's range or power requirements. They can include defective batteries (with defects or flaws) produced during the production process and waste batteries generated during testing.

[0076] Among them, the sample parameters of retired batteries refer to the relevant performance parameters of retired batteries, such as: basic parameters nominal voltage, nominal capacity, rated capacity, electrode system type, production batch, factory date, usage history, remaining capacity (SOH) internal resistance, state of charge (SOC), open circuit voltage (OCV), number of cycles, charge and discharge efficiency (coulombic efficiency), etc., which are used to assist in the screening of retired batteries. Performance and production parameters.

[0077] Step S20: constructing an incremental capacity curve according to the sample parameters of the retired battery, and extracting a curve peak feature based on the incremental capacity curve.

[0078] It should be noted that grouping batteries with inconsistent performance into a single unit for subsequent recycling can negatively impact the overall performance of the battery module and even create safety risks during subsequent sorting and recycling. To address this issue, the proposed method considers the static performance of retired batteries and screens out batteries with inconsistent static parameters, effectively improving sorting efficiency and safety.

[0079] Furthermore, from the perspective of the operating mechanism of lithium-ion batteries, the capacity decay process of lithium-ion batteries is related to the loss of lithium ions during the charge and discharge process, the formation of a solid electrolyte interface layer, and electrolyte degradation. These processes all lead to the obstruction of ion transport and the reduction of electron conduction efficiency, which manifests itself macroscopically as an increase in the internal resistance of the lithium-ion battery. Therefore, the impedance of the lithium-ion battery is correlated with the battery's aging state.

[0080] In its specific implementation, this solution uses the incremental capacity analysis (ICA) method to extract operating characteristics related to the health status of lithium batteries. During the charging process of a lithium battery, the voltage rises slowly in the middle of charging, making it difficult to observe the charging process of the lithium battery through voltage changes. However, most of the capacity of the lithium battery is charged in the middle of charging, and the characteristics of the lithium battery in the middle of charging are closely related to its health status. Therefore, this solution uses the ICA method to analyze the voltage charging process. By differentially processing the capacity and voltage of the lithium battery during the charging process, the difficult-to-observe lithium battery voltage changes are converted into an IC curve to determine the performance differences of lithium batteries in different health states and extract the peak characteristics of the IC curve. In summary, the two major performance parameters, resistance and IC curve peak, can well reflect the aging process of lithium batteries. This solution extracts the above operating characteristics from the battery operation process to evaluate battery performance consistency and perform screening.

[0081] In some embodiments, the specific incremental capacity curve construction process may be to perform incremental capacity analysis based on the charging process capacity and voltage in the sample parameters to obtain initial incremental capacity data; and to filter the initial incremental capacity data using a filtering algorithm to construct the incremental capacity curve; the filtering algorithm may include:

[0082] ;

[0083] in is the output after filtering, is the first data points, are the coefficients obtained by polynomial fitting, is the window size.

[0084] It should be noted that in order to ensure the stability of the extracted static parameters, this solution uses a Savitzky-Golay filter to process the data and extract a reliable IC curve. The Savitzky-Golay filter is a signal smoothing tool that improves data quality by fitting a polynomial to the data points in a sliding window and then replacing the original data points with the values ​​obtained by the fitted polynomial. The filter has two key parameters: the window size and the polynomial order. By adjusting these parameters, the degree of smoothing can be flexibly adjusted. This solution selects the portion of the incremental capacity curve corresponding to a voltage greater than 2.9 V and less than 3.3 V as input, thereby retaining key information about the peak portion of the IC curve while alleviating the problem that smaller signal values ​​are more easily amplified by noise during the derivative calculation process. The filter formula is expressed as follows:

[0085] ;

[0086] in is the output after filtering, is the first data points, are the coefficients obtained by polynomial fitting, is the window size.

[0087] Step S30: performing battery performance consistency screening on the retired batteries according to the peak characteristics of the curve and the corresponding battery internal resistance to obtain screened target retired batteries.

[0088] It should be noted that the consistency screening process for retired batteries based on incremental capacity (IC) curves can be summarized as "quantifying battery performance differences using electrochemical characteristics and achieving efficient sorting through multi-dimensional parameter fusion." Specifically, the process first generates IC curves by differentially processing the charge-discharge voltage-capacity curves. For retired batteries, aging can lead to IC peak area decay (reflecting active lithium loss), peak position shift (degradation of the electrode material structure), and peak shape broadening (increased polarization). These characteristics constitute quantitative indicators of battery performance degradation. During the sorting stage, key peak parameters of the IC curves, such as peak voltage, peak area, and full width at half maximum (FWHM), must first be extracted. By establishing a mapping model between IC peak parameters and internal resistance (e.g., principal component analysis or support vector machine), retired batteries can be divided into highly consistent clusters (e.g., IC peak area variation <5%, internal resistance dispersion <10%) and abnormal battery clusters.

[0089] In some embodiments, the specific screening process can be determined based on cluster analysis. For example, a screening parameter set is generated based on the peak coordinates in the peak characteristics of the curve and the internal resistance of the battery; a cluster analysis is performed on the screening parameter set based on a preset area, a preset neighborhood radius, and a preset minimum density to determine core points, edge points, or noise points; the preset area is: ,in, is the sample point and The Euclidean distance between them, Eps is the preset neighborhood radius; the retired batteries are screened for battery performance consistency according to the core points, edge points or noise points.

[0090] It should be noted that this solution uses the extracted static parameters as the input of the battery screening algorithm. DBSCAN achieves clustering by classifying each point as a core point, an edge point, or a noise point. Specifically, DBSCAN requires two initial parameters: neighborhood radius (Eps) and minimum density (MinPts). Eps represents the radius value used to determine the neighborhood size, and MinPts represents the minimum number of points that must exist within the Eps radius when a point is classified as a core point. Sample points in , whose neighborhood is defined as: . is the sample point and The Euclidean distance between them is calculated as follows:

[0091] ;

[0092] in is the dimension of the space, is the sample point Another point in the neighborhood.

[0093] If the sample point satisfy , it is classified as a core point. If a sample point is not a core point but is within the neighborhood of a core point, it is considered directly density-reachable from the core point and classified as an edge point. If a point is directly density-reachable from a core point or edge point in at least one cluster, it is also density-connected to other points in that cluster and is included in that cluster. Points that do not belong to any cluster are classified as noise points. Iterate the neighborhood of each core point and perform a minimum density check until all battery points have been visited.

[0094] This solution selects the peak horizontal and vertical coordinates of the IC curve and the battery's internal resistance as the parameter set for battery consistency screening. These three parameters are input into the model in pairs, and the screening results are cross-validated to effectively screen out retired batteries with inconsistent performance.

[0095] Step S40: performing feature extraction based on the sample parameters of the target retired battery to determine a battery dynamic performance feature image and a battery operating status feature image.

[0096] It should be noted that the feature image is extracted because the final classification can be calculated through the classification model. By inputting the feature image to classify the image, it is determined which group the target retired battery belongs to, and the classification of the retired battery is completed. Specifically, the battery dynamic performance feature image and the battery operating status feature image are images that represent the battery operating status and battery dynamic performance, respectively. For example, the image of the IC curve (Incremental Capacity) and the image of the VQ curve (voltage-capacity curve). The VQ curve directly reflects the relationship between the battery's charge and discharge capacity and the voltage platform. Its slope change can represent dynamic processes such as electrolyte decomposition and interfacial impedance growth. For example, the flat voltage platform of lithium iron phosphate (LFP) batteries is significantly different from the slope curve of nickel manganese cobalt oxide (NMC).

[0097] In some embodiments, the image acquisition process of the battery dynamic performance can be as follows: determining a voltage-capacity change curve based on sample parameters of the target retired battery; normalizing the voltage-capacity change curve to obtain a standardized voltage-capacity change curve; and interpolating the standardized voltage-capacity change curve to generate a battery operating status characteristic image of a preset size.

[0098] It should be noted that the static performance of a battery is affected by various factors, including temperature, changes in charge and discharge state, and capacity regeneration. Therefore, classifying batteries based solely on static performance parameters may lead to misclassification. Therefore, this solution introduces battery operating data to characterize its dynamic performance. First, the VQ curve is introduced to represent the relationship between voltage and capacity during charging and discharging. Generally, healthy batteries have a smoother voltage-to-capacity curve, while aging batteries may experience voltage fluctuations or capacity drops. The plateau voltage in the VQ curve is also an important indicator of battery health.

[0099] In addition, this scheme combines IC curves to analyze the changes in the battery's response ability to unit voltage changes. The IC curve can provide more detailed microscopic information for the classification model, such as the degradation of electrode materials and the aging of electrolytes, thereby having a more comprehensive understanding of the dynamic performance of the battery.

[0100] In the specific implementation, after extracting the dynamic performance data of the battery, in order to fully preserve the long-term dependencies of the data, this solution uses a data imaging method to convert the curve into a grayscale image. By retaining the overall trend and local features of the data, it provides richer information for the classification model and realizes a comprehensive characterization of the dynamic performance.

[0101] Specifically, in order to preserve the overall operating information of the battery, the VQ curve is converted into a grayscale image through interpolation and grayscale conversion. Assume that a VQ curve sequence is , first normalized to achieve a standard representation of the data.

[0102] ;

[0103] in for The data in represents the normalized data, and Represent the maximum and minimum values ​​of the sequence respectively. Next, all data are normalized to a uniform sequence length using interpolation. To ensure that the subsequent image size is consistent. In order to align the range of pixel values ​​of each pixel in the data and the grayscale image, this solution scales the data from the range [0, 1] to [0, 255]. Finally, it is generated by column replication. The grayscale image of the VQ curve is completely retained, and the input requirements of the subsequent image processing model are met, thereby obtaining the image.

[0104] In some embodiments, the image acquisition process may include performing incremental capacity analysis based on sample parameters of the target retired battery to extract an incremental capacity curve of the target retired battery; performing polar coordinate conversion on the incremental capacity curve to obtain a target polar coordinate sequence; calculating a vector inner product based on the target polar coordinate sequence to obtain a Gram matrix; and calculating a Gram angle sum field and a Gram angle difference field based on the target polar coordinate sequence. The calculation scheme includes:

[0105] ;

[0106] ;

[0107] in, is a unit vector, is the target polar coordinate sequence, φi and φj are the corresponding angles in polar coordinates; represents the matrix after Gram angle and field transformation, represents the matrix after Gram angle difference field transformation;

[0108] A battery dynamic performance characteristic image is determined according to the Gram matrix, the Gram angle sum field, and the Gram angle difference field.

[0109] In the specific implementation, this scheme uses incremental capacity analysis to extract IC curves, thereby enhancing the characterization of battery dynamic performance. For the obtained IC curves, GAF is used to encode the one-dimensional sequence into a two-dimensional image while retaining the time correlation and position information of the data. The process of mapping the sequence to the GAF image involves two main steps: polar coordinate transformation and GAF ​​transformation. First, the time series is normalized, and then each data point is represented as an angle and radius through polar coordinate transformation. Next, in the polar coordinate system, the Gramian matrix (GM) is generated by calculating the trigonometric function of the sum or difference of the angles between the data points. GAF uses the polar coordinate system to represent the sequence, encodes the time position information as the radius, and the value of the sequence is the angle. It performs an inverse cosine calculation on the scaled value to obtain the corresponding angle in the polar coordinate. , and encode the timestamp as the radius , as shown below:

[0110] ;

[0111] in and Represent the angle and radius of polar coordinate transformation, thereby preserving the amplitude and time dependence between sequence data. Indicates the The timestamp of the location, Indicates the length of the sequence. It is worth noting that within the range, is monotonic, so for a given time series in polar coordinates, the generated result is unique. Therefore, the entire image encoding process is a bidirectional mapping without losing any information about the IC curve. The Gram matrix consists of vector inner products that capture the angular and directional relationships between similar sequences. For a one-dimensional sequence , GM is defined as:

[0112] ;

[0113] in represents the generated Gram matrix, =X1 X2 COS , is the angle between two vectors. The generated GM is converted to a grayscale image, with each pixel value corresponding to an element of the matrix, thus preserving the temporal dependence and dynamic characteristics of the time series. To further extract the overall trend and peak characteristics of the IC curve, this scheme introduces the Gramian Angular Summation Field (GASF) and the Gramian Angular Difference Field (GADF). These fields characterize the correlation of sequential data by summing or differencing trigonometric functions. Although the images produced by GASF and GADF are similar to those produced by GAF, GASF emphasizes the overall relationship between data points by applying the sum of angles. GADF, on the other hand, generates the GM by calculating the angular differences between data points, producing an image that highlights the variation in the data. In summary, GASF captures global trends, while GADF emphasizes local variations. In this scheme, both GASF and GADF are used to extract features from the IC curve. GASF effectively captures the overall trend of the curve, while GADF is better at extracting transient features. The calculations of GASF and GADF are as follows:

[0114] ;

[0115] ;

[0116] in is a unit vector, is the normalized sequence.

[0117] Finally, this scheme uses piecewise aggregate approximation (PAA) to resize the generated GASF and GADF images. To facilitate subsequent experiments, all images were resized to 224×224. PAA is a technique for dimensionality reduction of sequence data. It partitions the original time series data into multiple subintervals and then simplifies the data by taking the average value of each interval. This reduces data complexity while preserving key features.

[0118] Step S50: inputting the battery dynamic performance characteristic image and the battery operating status characteristic image into an image classification model to complete battery screening based on dynamic and static characteristics.

[0119] It should be noted that the final screening with the help of the model requires the normalization of the input image, and the battery dynamic performance characteristic image and the battery operation status characteristic image are input as input to the image classification model. The specific classification process can be to use convolutional neural network (CNN) or visual transformer (ViT) as the backbone network combined with convolution kernels of different receptive fields, while capturing the features of static images, and dividing the battery levels (such as qualified products, maintenance-required products, and obsolete products) according to the preset confidence threshold (such as 0.8) or cluster analysis, and output the screening results to the sorting system for battery pack matching, fault warning or life prediction.

[0120] In its specific implementation, the overall architecture of this solution is screening-graphical representation-model-based feature image classification. First, density-based clustering is used to screen out batteries with abnormal performance. Then, data imaging and deep learning are used to graphically represent the dynamic performance of retired batteries, thereby achieving accurate battery sorting. The proposed method includes three steps: designing a battery screening method based on density clustering, using the peak coordinates of the internal resistance and incremental capacity curves as the performance parameters of the input model, screening out batteries with inconsistent performance based on DBSCAN, using grayscale transformation and Gram angle field transformation of the voltage-capacity curve and incremental capacity curve, converting the battery dynamic performance data into images, and achieving accurate characterization of dynamic performance. Using three-channel features as input, a densely connected convolutional neural network combined with an attention mechanism is used for feature extraction and representation, improving the recognition ability of dynamic feature images and achieving accurate classification.

[0121] This embodiment obtains sample parameters of retired batteries; constructs an incremental capacity curve based on the sample parameters of the retired batteries, and extracts the peak characteristics of the curve based on the incremental capacity curve; performs battery performance consistency screening on the retired batteries based on the peak characteristics of the curve and the corresponding battery internal resistance to obtain the screened target retired batteries; performs feature extraction based on the sample parameters of the target retired batteries to determine the battery dynamic performance feature image and the battery operating status feature image; inputs the battery dynamic performance feature image and the battery operating status feature image into the image classification model to complete battery screening based on dynamic and static features. The above method realizes the consistency screening of retired batteries, thereby providing a good data foundation for battery classification and improving the accuracy of retired battery screening.

[0122] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above embodiment 1 can be referred to the above introduction and will not be described in detail later. Figure 2 In step S50, the battery screening method based on dynamic and static characteristics further includes:

[0123] Step S51: The battery dynamic performance characteristic image and the battery operation status characteristic image are used in a backbone network to determine the high-dimensional vector features of the target retired battery. The backbone network includes a densely connected network and a CBAM feature extraction and optimization module.

[0124] This network utilizes a dual-path feature fusion framework, with its core consisting of a densely connected network (DenseNet) and a CBAM optimization module. This network aims to extract high-dimensional vectors of both dynamic and static battery features through the synergistic reuse of dense features and an attention mechanism. It includes an input layer that receives images of the battery's operating status (static features) and dynamic performance images (dynamic features such as recurrence graphs and incremental capacity curves), resized to a fixed size (e.g., 224×224 pixels). The dense block and CBAM modules extract basic features through multiple dense blocks, and a CBAM module is embedded after each dense block for feature optimization. The transition layer performs compression: 1×1 convolutions are used to reduce the number of channels, combined with average pooling to compress spatial dimensions and prevent feature redundancy. Finally, global average pooling is used to generate a high-dimensional vector for subsequent classification or regression tasks. The Convolutional Block Attention Module (CBAM) is a lightweight module that combines channel-wise and spatial-attention, aiming to improve the performance of convolutional neural networks through adaptive feature refinement. The core idea is to dynamically adjust the response strength of key areas by learning the importance weights of different channels and spatial positions in the feature map, thereby enhancing the model's ability to capture key information.

[0125] In some embodiments, the network structure of the densely connected network is characterized by:

[0126] ,in, Is a comprehensive function that includes multiple operations, Indicates the 0th to The output, For the The output of the layer; the densely connected network includes: ReLu activation function and 3x3 convolution; the formulas of the channel attention mechanism and spatial attention mechanism of the CBAM feature extraction optimization module are:

[0127] ;

[0128] in, represents global average pooling, represents the maximum pooling operation, represents a multilayer perceptron, represents the Sigmoid activation function, Represents a convolution operation.

[0129] Step S52: Input the high-dimensional vector features into the classification layer to determine the category probability.

[0130] It should be noted that class probabilities can be obtained by, for example, first performing a linear transformation on the high-dimensional feature vector through a fully connected layer, mapping the feature dimensions to the class space. For example, if the classification task has four classes, the fully connected layer will convert the input high-dimensional features (e.g., 12 dimensions) into a 4-dimensional output, with each dimension corresponding to a class score. The raw scores are then converted into probability distributions using an activation function, such as the Softmax or Sigmoid function.

[0131] Step S53: completing battery screening based on dynamic and static features according to the category probability.

[0132] It is understandable that the results of the consistency screening of retired batteries in this embodiment are illustrated by the following cases. For example, in the battery screening process using the IC curve peak voltage and internal resistance as input, batteries with too large or too small internal resistance are difficult to cluster into groups with other batteries and can be effectively screened out. In the battery screening process using the IC curve peak horizontal and vertical coordinates as input, batteries that are obviously inconsistent with other battery performance are effectively screened out. In the battery screening process using the IC curve peak vertical coordinate and internal resistance as input, batteries with too low or too high internal resistance and poor consistency with other batteries are effectively screened out.

[0133] In this embodiment, the backbone network of the battery dynamic performance characteristic image and the battery operating status characteristic image determines the high-dimensional vector features of the target retired battery. The backbone network includes a densely connected network and a CBAM feature extraction and optimization module; the high-dimensional vector features are input into the classification layer to determine the category probability; and the battery screening based on dynamic and static features is completed according to the category probability. This method enhances feature reusability by fusing features through a densely connected network and a CBAM attention module, strengthens the weights of key dynamic parameters through channel and spatial attention mechanisms, and ultimately achieves collaborative optimization of dynamic and static consistency. After screening, the voltage response differences of the battery pack are significantly reduced, the capacity distribution is compact, and the cascade utilization efficiency and safety are simultaneously improved.

[0134] In addition, an embodiment of the present invention further proposes a storage medium on which a battery screening program based on dynamic and static features is stored. When the battery screening program based on dynamic and static features is executed by a processor, the steps of the battery screening method based on dynamic and static features as described above are implemented.

[0135] Reference Figure 3 , Figure 3 This is a structural block diagram of the first embodiment of the battery screening device based on dynamic and static characteristics of the present invention.

[0136] like Figure 3 As shown, the battery screening device based on dynamic and static characteristics proposed in the embodiment of the present invention includes:

[0137] An acquisition module 10 is used to obtain sample parameters of retired batteries;

[0138] a processing module 20 for constructing an incremental capacity curve according to the sample parameters of the retired battery, and extracting a curve peak feature based on the incremental capacity curve;

[0139] The processing module 20 is configured to perform battery performance consistency screening on the retired batteries according to the peak characteristics of the curve and the corresponding battery internal resistance, to obtain screened target retired batteries;

[0140] The processing module 20 is configured to extract features based on the sample parameters of the target retired battery to determine a battery dynamic performance feature image and a battery operating status feature image;

[0141] The processing module 20 is used to input the battery dynamic performance characteristic image and the battery operation status characteristic image into the image classification model to complete the battery screening based on dynamic and static characteristics.

[0142] The battery screening device based on dynamic and static features provided in this application adopts the battery screening method based on dynamic and static features in the above-mentioned embodiments, and can solve the technical problems of battery screening based on dynamic and static features. Compared with the existing technology, the beneficial effects of the battery screening device based on dynamic and static features provided in this application are the same as the beneficial effects of the battery screening method based on dynamic and static features provided in the above-mentioned embodiments, and the other technical features of the battery screening device based on dynamic and static features are the same as the features disclosed in the above-mentioned embodiment methods, and are not further described here.

[0143] It should be understood that the above is only an example and does not constitute any limitation to the technical solution of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any limitation on this.

[0144] In this embodiment, the acquisition module 10 acquires sample parameters of retired batteries; the processing module 20 constructs an incremental capacity curve based on the sample parameters of the retired batteries, and extracts the peak characteristics of the curve based on the incremental capacity curve; the processing module 20 performs battery performance consistency screening on the retired batteries based on the peak characteristics of the curve and the corresponding battery internal resistance to obtain the screened target retired batteries; the processing module 20 performs feature extraction based on the sample parameters of the target retired batteries to determine the battery dynamic performance feature image and the battery operating status feature image; the processing module 20 inputs the battery dynamic performance feature image and the battery operating status feature image into the image classification model to complete battery screening based on dynamic and static features. The above method realizes the consistency screening of retired batteries, thereby providing a good data foundation for battery classification and improving the accuracy of retired battery screening.

[0145] In some embodiments, the processing module 20 is configured to perform incremental capacity analysis based on the charging process capacity and voltage in the sample parameters to obtain initial incremental capacity data;

[0146] The initial incremental capacity data is filtered by a filtering algorithm to construct an incremental capacity curve;

[0147] The filtering algorithm includes:

[0148] ;

[0149] in is the output after filtering, is the first data points, are the coefficients obtained by polynomial fitting, is the window size.

[0150] In some embodiments, the processing module 20 is configured to generate a screening parameter set based on the peak coordinates in the peak feature of the curve and the internal resistance of the battery;

[0151] Performing cluster analysis on the screening parameter set based on a preset domain, a preset neighborhood radius, and a preset minimum density to determine core points, edge points, or noise points;

[0152] The preset fields are: ,in, is the sample point and The Euclidean distance between them, Eps is the preset neighborhood radius;

[0153] The retired batteries are screened for battery performance consistency based on the core points, edge points or noise points.

[0154] In some embodiments, the processing module 20 is configured to determine a voltage-capacity variation curve based on sample parameters of the target retired battery;

[0155] Normalizing the voltage-capacity change curve to obtain a standardized voltage-capacity change curve;

[0156] Interpolation processing is performed on the standardized voltage-capacity change curve to generate a battery operation status characteristic image of a preset size.

[0157] In some embodiments, the processing module 20 is configured to perform incremental capacity analysis based on the sample parameters of the target retired battery to extract an incremental capacity curve of the target retired battery;

[0158] Performing polar coordinate conversion on the incremental capacity curve to obtain a target polar coordinate sequence;

[0159] Calculating the inner product of the vectors according to the target polar coordinate sequence to obtain a Gram matrix;

[0160] Calculating the Gram angle sum field and the Gram angle difference field based on the target polar coordinate sequence, the calculation scheme includes:

[0161] ;

[0162] ;

[0163] in, is a unit vector, is the target polar coordinate sequence, φi and φj are the corresponding angles in polar coordinates; represents the matrix after Gram angle and field transformation, represents the matrix after Gram angle difference field transformation;

[0164] A battery dynamic performance characteristic image is determined according to the Gram matrix, the Gram angle sum field, and the Gram angle difference field.

[0165] In some embodiments, the processing module 20 is configured to determine high-dimensional vector features of a target retired battery using a backbone network of the battery dynamic performance characteristic image and the battery operating status characteristic image, wherein the backbone network includes a densely connected network and a CBAM feature extraction and optimization module;

[0166] Inputting the high-dimensional vector features into a classification layer to determine the category probability;

[0167] Battery screening based on dynamic and static features is completed according to the category probability.

[0168] The present application provides a battery screening device based on dynamic and static features, and the battery screening device based on dynamic and static features includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the battery screening method based on dynamic and static features in the above-mentioned embodiment one.

[0169] Reference below Figure 4 , which shows a schematic diagram of the structure of a battery screening device based on dynamic and static features suitable for implementing an embodiment of the present application. The battery screening device based on dynamic and static features in the embodiment of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The battery screening device based on dynamic and static characteristics shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0170] like Figure 4As shown, the battery screening device based on dynamic and static characteristics may include a processing device 1001 (such as a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in the ROM (Read Only Memory) 1002 or the program loaded from the storage device 1003 to the RAM (Random Access Memory) 1004. Various programs and data required for the operation of the battery screening device based on dynamic and static characteristics are also stored in the RAM 1004. The processing device 1001, ROM 1002 and RAM 1004 are connected to each other via a bus 1005. The input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the battery screening device based on dynamic and static characteristics to communicate wirelessly or wired with other devices to exchange data. Although the figure shows a battery screening device based on dynamic and static characteristics with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have instead.

[0171] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0172] The battery screening device based on dynamic and static features provided in this application adopts the battery screening method based on dynamic and static features in the above-mentioned embodiment, which can solve the technical problems of battery screening based on dynamic and static features. Compared with the existing technology, the beneficial effects of the battery screening device based on dynamic and static features provided in this application are the same as the beneficial effects of the battery screening method based on dynamic and static features provided in the above-mentioned embodiment, and the other technical features of the battery screening device based on dynamic and static features are the same as the features disclosed in the method of the previous embodiment, and will not be repeated here.

[0173] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0174] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0175] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, a computer program) stored thereon, wherein the computer-readable program instructions are used to execute the battery screening method based on dynamic and static characteristics in the above-mentioned embodiment.

[0176] The computer-readable storage medium provided herein may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems, or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including, but not limited to, wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0177] The computer-readable storage medium may be included in the battery screening device based on dynamic and static characteristics; or it may exist independently without being assembled into the battery screening device based on dynamic and static characteristics.

[0178] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the battery screening device based on dynamic and static features, the battery screening device based on dynamic and static features: obtains sample parameters of retired batteries; constructs an incremental capacity curve based on the sample parameters of the retired batteries, and extracts the peak characteristics of the curve based on the incremental capacity curve; performs battery performance consistency screening on the retired batteries based on the peak characteristics of the curve and the corresponding battery internal resistance to obtain the screened target retired batteries; performs feature extraction based on the sample parameters of the target retired batteries to determine the battery dynamic performance feature image and the battery operating status feature image; inputs the battery dynamic performance feature image and the battery operating status feature image into the image classification model to complete battery screening based on dynamic and static features. The above method realizes the consistency screening of retired batteries, thereby providing a good data foundation for battery classification and improving the accuracy of retired battery screening.

[0179] Computer program code for performing the operations of the present application may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0180] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, program segment or a part of code, and the module, program segment or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented by a dedicated hardware-based system that performs the specified function or operation, or can be implemented by a combination of dedicated hardware and computer instructions.

[0181] The modules described in the embodiments of the present application may be implemented in software or hardware, wherein the name of a module does not necessarily limit the unit itself.

[0182] The computer-readable storage medium provided in this application stores computer-readable program instructions (i.e., a computer program) for executing the aforementioned method for battery screening based on dynamic and static characteristics. This computer-readable storage medium can address the technical issues surrounding battery screening based on dynamic and static characteristics. Compared to the prior art, the beneficial effects of the computer-readable storage medium provided in this application are similar to those of the battery screening method based on dynamic and static characteristics provided in the aforementioned embodiments, and are not further elaborated here.

[0183] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned battery screening method based on dynamic and static characteristics when executed by a processor.

[0184] The computer program product provided in this application can solve the technical problem of battery screening based on dynamic and static characteristics. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the battery screening method based on dynamic and static characteristics provided in the above embodiment, and will not be repeated here.

[0185] The above description is only part of the embodiments of the present application and does not limit the patent scope of the present application. All equivalent structural transformations made by using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A battery screening method based on dynamic and static characteristics, characterized in that: The battery screening method based on dynamic and static characteristics includes: Obtain sample parameters of retired batteries; constructing an incremental capacity curve according to the sample parameters of the retired battery, and extracting a curve peak feature based on the incremental capacity curve; Performing battery performance consistency screening on the retired batteries according to the peak characteristics of the curve and the corresponding battery internal resistance to obtain screened target retired batteries; Extract features based on sample parameters of the target retired battery to determine a battery dynamic performance feature image and a battery operating status feature image; Inputting the battery dynamic performance characteristic image and the battery operating status characteristic image into an image classification model to complete battery screening based on dynamic and static characteristics; The step of extracting features based on sample parameters of the target retired battery to determine a battery operating status feature image includes: Determining a voltage-capacity variation curve according to sample parameters of the target retired battery; Normalizing the voltage-capacity change curve to obtain a standardized voltage-capacity change curve; Interpolation processing is performed on the standardized voltage-capacity change curve to generate a battery operation status characteristic image of a preset size.

2. The method according to claim 1, wherein The constructing an incremental capacity curve according to the sample parameters of the retired battery includes: Performing incremental capacity analysis based on the charging process capacity and voltage in the sample parameters to obtain initial incremental capacity data; The initial incremental capacity data is filtered by a filtering algorithm to construct an incremental capacity curve; The filtering algorithm includes: ； in is the output after filtering, is the first data points, are the coefficients obtained by polynomial fitting, is the window size.

3. The method according to claim 1, wherein The performing battery performance consistency screening on the retired battery according to the peak characteristics of the curve and the corresponding battery internal resistance includes: Generate a screening parameter set according to the peak coordinates in the peak characteristics of the curve and the internal resistance of the battery; Performing cluster analysis on the screening parameter set based on a preset domain, a preset neighborhood radius, and a preset minimum density to determine core points, edge points, or noise points; The preset fields are: ,in, is the sample point and The Euclidean distance between them, Eps is the preset neighborhood radius; The retired batteries are screened for battery performance consistency based on the core points, edge points or noise points.

4. The method according to claim 1, wherein The extracting features based on the sample parameters of the target retired battery to determine the battery dynamic performance characteristic image includes: Performing incremental capacity analysis on the sample parameters of the target retired battery to extract an incremental capacity curve of the target retired battery; Performing polar coordinate conversion on the incremental capacity curve to obtain a target polar coordinate sequence; Calculating the inner product of the vectors according to the target polar coordinate sequence to obtain a Gram matrix; Calculating the Gram angle sum field and the Gram angle difference field based on the target polar coordinate sequence, the calculation scheme includes: ； ； in, is a unit vector, is the target polar coordinate sequence, φi and φj are the corresponding angles in polar coordinates, represents the matrix after Gram angle and field transformation, represents the matrix after Gram angle difference field transformation; A battery dynamic performance characteristic image is determined according to the Gram matrix, the Gram angle sum field, and the Gram angle difference field.

5. The method according to claim 1, wherein Inputting the battery dynamic performance characteristic image and the battery operating status characteristic image into the image classification model to complete battery screening based on dynamic and static characteristics includes: Determine the high-dimensional vector features of the target retired battery based on the battery dynamic performance characteristic image and the battery operating status characteristic image backbone network, wherein the backbone network includes a dense connection network and a CBAM feature extraction and optimization module; Inputting the high-dimensional vector features into a classification layer to determine the category probability; Battery screening based on dynamic and static features is completed according to the category probability.

6. The method according to claim 5, wherein The network structure of the densely connected network is characterized as follows: ； in, Is a comprehensive function that includes multiple operations, Indicates the 0th to The output, For the The output of the layer; The densely connected network includes: ReLu activation function and 3x3 convolution; The formulas for the channel attention mechanism and spatial attention mechanism of the CBAM feature extraction and optimization module are: ； ； in, represents global average pooling, represents the maximum pooling operation, represents a multilayer perceptron, represents the Sigmoid activation function, Represents a convolution operation.

7. A battery screening device based on dynamic and static characteristics, characterized in that: The method according to claim 1 is performed, wherein the battery screening device based on dynamic and static characteristics comprises: An acquisition module, used to obtain sample parameters of retired batteries; a processing module, configured to construct an incremental capacity curve according to sample parameters of the retired battery, and extract a curve peak feature based on the incremental capacity curve; The processing module is configured to perform battery performance consistency screening on the retired batteries according to the peak characteristics of the curve and the corresponding battery internal resistance, to obtain screened target retired batteries; The processing module is configured to extract features based on sample parameters of the target retired battery to determine a battery dynamic performance feature image and a battery operating status feature image; The processing module is used to input the battery dynamic performance characteristic image and the battery operation status characteristic image into the image classification model to complete the battery screening based on dynamic and static characteristics.

8. A battery screening device based on dynamic and static characteristics, characterized in that: The device includes: a memory, a processor, and a battery screening program based on dynamic and static features stored in the memory and executable on the processor, wherein the battery screening program based on dynamic and static features is configured to implement the steps of the battery screening method based on dynamic and static features as described in any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium stores a battery screening program based on dynamic and static features, and when the battery screening program based on dynamic and static features is executed by the processor, the steps of the battery screening method based on dynamic and static features according to any one of claims 1 to 6 are implemented.

Images