Lithium ion battery outlier early warning method, system, product, medium and equipment

Abstract

The invention discloses a lithium ion battery outlier early warning method and system, a product, a medium and equipment, and relates to the technical field of battery fault prediction. The method comprises the steps that electrical performance testing is carried out on the lithium ion battery, a life evaluation model comprising three branches is constructed, the life evaluation model is trained, and the three branches comprise a first branch used for local outlier detection, a second branch used for residual analysis and a third branch used for deep learning; the three branches in the life evaluation model process the direct current internal resistance, the constant voltage stage capacity ratio and the constant current stage capacity ratio in parallel, fusion decision is carried out on results obtained by the three branches respectively, and an outlier detection result is obtained. According to the method, the service life attenuation rule of the battery is calculated based on the capacity ratio in the constant-voltage stage, the advantages of actual experiments and model learning can be brought into full play, and accurate recognition and early warning of the lithium ion battery outlier phenomenon are achieved through comprehensive analysis of multi-modal data.

Description

Technical Field

[0001] The present invention relates to the technical field of battery failure prediction, and in particular to a lithium-ion battery outlier warning method, system, product, medium and equipment. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Lithium-ion batteries, as core energy carriers in numerous fields such as new energy vehicles, energy storage systems, and consumer electronics, are crucial for their performance and safety. However, over long-term use, lithium-ion batteries often exhibit outlier behavior, whereby the parameters of individual cells or modules significantly deviate from normal ranges. This can manifest as capacity fade, increased internal resistance, lithium deposition, thermal runaway, and other issues. These issues not only degrade overall device performance but can also lead to safety incidents, posing a serious threat to user safety and property.

[0004] While numerous studies have been conducted on the detection and analysis of outliers in lithium-ion batteries, some shortcomings remain. Some methods focus on monitoring a single parameter, such as determining whether a cell is an outlier based solely on changes in voltage or internal resistance. However, battery outliers are often the result of multiple factors, and a single parameter cannot fully and accurately reflect the true state of a cell, easily leading to misjudgments or missed detections. For example, judging solely by voltage changes may overlook other potential issues, such as increased internal resistance, resulting in the failure to promptly identify cells in the early stages of failure.

[0005] While existing technologies have attempted comprehensive multi-parameter analysis, most fail to fully integrate multimodal data. For example, while some methods consider multiple parameters such as voltage and internal resistance simultaneously, they fail to achieve simultaneous acquisition and in-depth coupled analysis of data from different parameters during data collection and analysis. This makes it difficult to accurately capture the characteristics of outlier cells when faced with complex battery state changes, reducing the accuracy and reliability of outlier detection.

[0006] Currently, single detection methods have limitations. For example, some methods that focus on experiments can cause significant damage to batteries during the detection process, while methods that rely on mathematical models are overly dependent on model accuracy and training data, resulting in varying detection efficiency. Furthermore, these methods often struggle to provide early warning of battery outliers, failing to provide accurate information to the battery management system in a timely manner to implement appropriate measures to prevent safety incidents.

[0007] In summary, existing technologies for detecting and analyzing lithium-ion battery outliers suffer from problems such as a single detection dimension, a single method, and difficulty achieving early warning. These issues limit the effective identification and management of lithium-ion battery outliers. A new method for comprehensive, accurate, and efficient battery outlier detection is urgently needed to improve the safety and reliability of battery systems, ensure the stable operation of related equipment, and ensure user safety. Summary of the Invention

[0008] In response to the shortcomings of the existing technology, the purpose of the present invention is to provide a lithium-ion battery outlier warning method, system, product, medium and equipment, which calculates the battery life attenuation law based on the capacity ratio in the constant voltage stage, and can give full play to the respective advantages of actual experiments and model learning. Through the comprehensive analysis of multimodal data, it can realize the accurate identification and warning of lithium-ion battery outlier phenomena.

[0009] In order to achieve the above object, the present invention is implemented through the following technical solutions: A first aspect of the present invention provides a lithium-ion battery outlier warning method, comprising the following steps: Conduct electrical performance tests on lithium-ion batteries to obtain the DC internal resistance, constant voltage stage capacity ratio, and constant current stage capacity ratio at different temperatures and different power levels; Constructing a lifespan assessment model comprising three branches and training the lifespan assessment model, wherein the three branches include a first branch for local outlier detection, a second branch for residual analysis, and a third branch for deep learning; The three branches in the life assessment model process the DC internal resistance, constant voltage stage capacity ratio, and constant current stage capacity ratio in parallel, and the results obtained from the three branches are integrated and decided to obtain the outlier detection results.

[0010] Furthermore, the first branch performs local outlier detection by constructing a matrix diagram, the second branch performs residual analysis by constructing a fitting formula based on the linear attenuation law and sinusoidal fluctuations, and the third branch captures the life attenuation law of the lithium battery according to the SOC DC internal resistance, the capacity ratio in the constant voltage stage, and the capacity ratio in the constant current stage to obtain the outlier prediction results of the lithium battery.

[0011] Furthermore, the third branch uses a multi-head attention mechanism to learn the relationship between SOC DC internal resistance, constant voltage stage capacity ratio, constant current stage capacity ratio and lithium battery life attenuation, and adopts a correction mechanism to correct the learned life attenuation law.

[0012] Furthermore, the specific steps for conducting electrical performance testing on lithium-ion batteries are as follows: Set different current parameters, voltage parameters and temperature parameters; Carry out electrical performance test experiments on charging and discharging of the lithium-ion battery to be tested; The electrical performance test experiment was repeated by changing the current parameters, voltage parameters and temperature parameters to obtain different test results.

[0013] Furthermore, the training steps of the lifespan assessment model are: Obtain multiple sets of lithium-ion battery electrical performance test data to form a data set, preprocess the parameters in the data set, and divide the data into training set and test set after annotation; The lifespan assessment model is trained using the training set, and the training effect is tested using the test set to obtain the trained lifespan assessment model.

[0014] Furthermore, the specific steps for preprocessing the electrical performance parameters are as follows: Clean the electrical performance parameters to remove noise data and outliers; The electrical performance parameters of different dimensions and ranges are normalized so that they are in the same numerical range.

[0015] A second aspect of the present invention provides a lithium-ion battery outlier warning system, comprising: The data acquisition module is configured to perform electrical performance tests on the lithium-ion battery, including DC internal resistance at different temperatures and different power levels, capacity percentage in the constant voltage phase, and capacity percentage in the constant current phase; a model building module configured to build a life assessment model comprising three branches and train the life assessment model, wherein the three branches include a first branch for local outlier detection, a second branch for residual analysis, and a third branch for deep learning; The outlier prediction module is configured as the three branches in the life assessment model to process the DC internal resistance, constant voltage stage capacity ratio, and constant current stage capacity ratio in parallel, and integrate the results obtained from the three branches to make decisions and obtain the outlier detection results.

[0016] A third aspect of the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program is suitable for being loaded by a processor and executing the steps of the lithium-ion battery outlier warning method as described in the first aspect of the present invention.

[0017] A fourth aspect of the present invention provides a computer device, comprising: a processor adapted to execute a computer program; A computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the lithium-ion battery outlier warning method according to the first aspect of the present invention is implemented.

[0018] A fifth aspect of the present invention provides a computer program product or computer program, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of the lithium-ion battery outlier warning method described in the first aspect of the present invention.

[0019] One or more of the above technical solutions have the following beneficial effects: This application discloses a lithium-ion battery outlier warning method, system, product, medium, and device. By comprehensively analyzing multimodal data using multiple methods, this method accurately identifies and warns of outliers in lithium-ion batteries. This method not only considers the key parameter of direct current internal resistance (DCIR), but also incorporates multi-dimensional data such as the constant voltage phase capacity percentage and the constant current phase capacity percentage. By comprehensively considering how these parameters vary under different temperatures and charge states, outlier detection is more accurate.

[0020] The present invention can detect outlier features according to different outlier prediction methods, and integrate and analyze all the detection results. This method can give full play to the respective advantages of experimental detection and mathematical model detection, and is more adaptable. This adaptability not only improves the generalization ability of the model, but also ensures that the health status and life decay laws of the battery can be accurately evaluated under various complex working conditions. The present invention realizes the accurate identification and early warning of outlier phenomena in lithium-ion batteries. This method not only improves the accuracy of outlier detection, enhances the adaptability and learning ability of the model, but also provides comprehensive battery health assessment and early warning functions, significantly improving the intelligence level and reliability of the battery management system. These technical effects make the present invention have important application value and broad development prospects in the field of management and maintenance of lithium-ion batteries.

[0021] The present invention utilizes a multi-head attention mechanism to learn the relationships between the SOC DC internal resistance, the constant voltage phase capacity ratio, the constant current phase capacity ratio, and the lifespan decay of lithium batteries. The multi-head attention mechanism can automatically focus on important features and relationships in the data, improving the model's ability to learn complex data patterns. In addition, the learned lifespan decay patterns are corrected through a correction mechanism, further improving the accuracy and reliability of the model. This combination of deep learning and a correction mechanism enables the model to more accurately capture the battery lifespan decay patterns, thereby achieving more reliable outlier warnings.

[0022] By combining comprehensive analysis of multimodal data with the predictive power of deep learning models, the present invention can proactively identify cells that are likely to exhibit outliers. This early warning capability enables the battery management system to implement preventative maintenance measures before problems escalate, thus preventing safety incidents.

[0023] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] 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, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0025] Figure 1 This is a flow chart of the lithium-ion battery outlier warning method in Example 1 of the present invention. DETAILED DESCRIPTION

[0026] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0027] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations; The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0028] Example 1: Embodiment 1 of the present invention provides a lithium-ion battery outlier warning method. In order to explore the reasons such as capacity decay, internal resistance increase, and lithium deposition that may occur in actual use conditions of lithium-ion batteries, which may cause individual battery cells to outlier in the battery pack, an outlier analysis method based on multi-field coupling is proposed to predict and determine in advance whether there are any abnormalities in the battery cells, so as to warn of safety accidents that may occur later.

[0029] like Figure 1 As shown, the specific steps include: S1: Conduct electrical performance tests on lithium-ion batteries to obtain the DC internal resistance, constant voltage stage capacity ratio, and constant current stage capacity ratio at different temperatures and different power levels.

[0030] S101: Setting different current parameters, voltage parameters and temperature parameters.

[0031] S102: Performing a charge and discharge electrical performance test experiment on the lithium-ion battery to be tested.

[0032] S103: Repeat the electrical performance test experiment by changing the current parameters, voltage parameters and temperature parameters to obtain different test results.

[0033] In a specific embodiment, the prepared battery is subjected to an electrical performance test to obtain electrical performance test result information, which includes direct current internal resistance (DCIR) at different SOCs at different temperatures and constant voltage stage capacity ratio.

[0034] This embodiment tests the DC internal resistance of lithium-ion batteries at different SOCs and the capacity CV ratio at cycle life attenuation. Specifically, four lithium iron phosphate batteries are tested for DC internal resistance (DCIR) at different SOCs at different temperatures: Experiment 1: Set the temperature to 25℃.

[0035] 1. 1C constant current charge to 3.65V, constant voltage charge to 0.05C.

[0036] 2. Let it sit for 30 minutes.

[0037] 3. 1C constant current discharge for 60S.

[0038] 4. Adjust the capacity to 95% SOC 5. Let it sit for 30 minutes.

[0039] 6. 1C constant current discharge for 60S.

[0040] Repeat steps 4 to 6 to obtain DCIR data at 90%, 80%, 50%, 20%, 10%, and 5% SOC, as shown in Table 1. Table 1. Results of battery health status (SOH) / battery charge (SOC) at 25°C.

[0041]

[0042] Experiment 2: Set the temperature to 45℃.

[0043] 1. 1C constant current charging to 3.65V, constant voltage charging to 0.05C: 2. Let it sit for 30 minutes; 3. 1C constant current discharge for 60S; 4. Adjust the capacity to 95% SOC 5. Let it sit for 30 minutes; 6. 1C constant current discharge for 60S; Repeat steps 4 to 6 in sequence to obtain DCIR data at 90%, 80%, 50%, 20%, 10%, and 5% SOC, as shown in Table 2.

[0044] Table 2. Results of battery health status SOH / battery charge SOC at 45°C.

[0045]

[0046] Experiment 3: Set the temperature to -10℃.

[0047] 1. 1C constant current charging to 3.65V, constant voltage charging to 0.05C: 2. Let it sit for 30 minutes; 3. 1C constant current discharge for 60S; 4. Adjust the capacity to 95% SOC 5. Let it sit for 30 minutes; 6. 1C constant current discharge for 60S; Repeat steps 4 to 6 in sequence to obtain DCIR data at 90%, 80%, 50%, 20%, 10%, and 5% SOC, as shown in Table 3.

[0048] Table 3. Results of battery health status SOH / battery charge SOC at -10°C.

[0049]

[0050] The measurement results of this example show the DC internal resistance at different SOCs at different temperatures (25°C, 45°C, and -10°C). The corresponding battery health status percentages (SOCs) are obtained based on existing cycle life tests. Tables 1, 2, and 3 show that the DC internal resistance of a battery under actual operating conditions affects its healthy charging and service life.

[0051] S2: Construct a lifespan assessment model containing three branches and train the lifespan assessment model.

[0052] In a specific implementation, this embodiment designs a hybrid model (Hybrid Matrix-Fitting-Relation Model) for lithium-ion battery outlier feature identification based on a fusion of matrix diagrams, fitting formulas, and deep learning. The three branches include a first branch for local outlier detection, a second branch for residual analysis, and a third branch for deep learning. The first branch performs local outlier detection by constructing a matrix diagram, the second branch performs residual analysis by constructing a fitting formula based on linear attenuation and sinusoidal fluctuations, and the third branch captures the lifespan attenuation law of the lithium battery based on the SOC DC internal resistance, the constant voltage phase capacity ratio, and the constant current phase capacity ratio to obtain outlier prediction results for the lithium battery.

[0053] S201: First, create a matrix based on the three temperature measurement results from S1. The constant voltage percentage determines the internal impedance of the battery. The higher the battery impedance, the greater the temperature rise and the higher the risk factor. If the battery data within the battery pack is not in the data matrix or the CV capacity percentage suddenly increases, it can be inferred that the battery is abnormal.

[0054] Specifically, in the first branch, the DCIR values ​​for each SOC point at 25°C, 45°C, and -10°C are input to construct a 3×N matrix M∈R^{3×N} (where N is the number of SOC points). A 2-D CNN (3×3 convolution kernel) is used to extract local spatial correlations, and channel attention is used to emphasize differences in the temperature dimension. Finally, a global average pooling operation is performed to obtain a 128-dimensional vector, which is then passed through a fully connected layer to output the outlier probability.

[0055] S202: The second branch performs linear fitting on the obtained results and determines whether the battery has an outlier abnormality according to the fitting formula.

[0056] This embodiment constructs a fitting formula based on the linear attenuation law and sinusoidal fluctuations: .

[0057] Among them, the linear part: Used to describe the overall downward trend, the fluctuation part: Used to describe periodic fluctuations. a represents the initial y-intercept, b represents the linear attenuation slope, where a negative value indicates a decrease, c represents the amplitude, d represents the frequency, e represents the phase offset, and y represents the predicted result.

[0058] Specifically, the second branch structure is a spatial feature extractor + bidirectional temporal modeler (CNN + Bi-LSTM) structure, which combines the local spatial feature extraction capability of the convolutional neural network (CNN) with the temporal modeling capability of the bidirectional long short-term memory network (Bi-LSTM). The outlier probability is calculated by residual analysis. Its input is the number of cycles, the measured SOH value, and the predicted value obtained by fitting with the above fitting formula. The designed residual δ is the difference between the predicted value and the actual value. The obtained residual sequence δ(x) is input into the 1-D CNN + Bi-LSTM to capture the nonlinearity and temporal drift of the residual with the cycle; the residual distribution parameters are output, and the negative log-likelihood loss is calculated. The outlier probability is output using the 3σ rule. After training, the fitting formula of this embodiment can be written as: .

[0059] S203: The third branch learns the relationship between SOC DC internal resistance, constant voltage stage capacity ratio, constant current stage capacity ratio and lithium battery life attenuation through a multi-head attention mechanism, and uses a correction mechanism to correct the learned life attenuation law.

[0060] The third branch of the lifespan assessment model consists of a backbone network, a deep learning network, and a classifier. The number of cycles is set, and the SOC DC internal resistance, constant voltage phase capacity ratio, and constant current phase capacity ratio at different temperatures are used as input. The backbone network is used to extract the different features of the input data and map the different dimensional features to a unified dimension. The deep learning network based on the multi-head attention mechanism learns the relationships between the features extracted by the backbone network. The learned results are classified by the classifier to obtain outlier prediction results.

[0061] Specifically, multi-head attention (8 heads) learns cross-feature coupling, and its formula is: .

[0062] Here, Query, Key, and Value represent Q, K, and V in the attention mechanism, respectively. f() represents the mapping function. SOC-DCIRall represents the DC internal resistance sequence at different SOC points. CV% and CC% represent the capacity percentages during the constant voltage phase and the constant current phase, respectively. A correction mechanism is used to suppress attention noise. Finally, the deep learning network outputs a 256-dimensional relationship vector, which is passed through a classifier to determine the outlier probability.

[0063] The correction mechanism involves estimating the noise in the multi-head attention mechanism, filtering it with learnable gating, and finally, after residual and normalization operations, outputting a stable 256-dimensional relationship vector. This correction mechanism uses the noise removal method of the existing attention mechanism and will not be detailed here.

[0064] S204: The three branches output the outlier probability or feature vector of the same battery sample, and finally give a comprehensive outlier score through the adaptive weight fusion module (AWF).

[0065] S2041: Set the outlier threshold θ and calculate the confidence of each branch.

[0066] S2042: Softmax normalization is used to obtain the confidence weights of different branches.

[0067] S2043: Calculate the composite outlier score: .

[0068] Where Sout is the comprehensive outlier score, w1, w2, and w3 are the confidence weights of the first, second, and third branches, respectively, and p1, p2, and p3 are the outputs of the first, second, and third branches, respectively. An outlier alarm is triggered when Sout ≥ θ.

[0069] S205: Train the H-MFR model using historical data or existing public data sets.

[0070] S2051: Acquire multiple sets of lithium-ion battery electrical performance test data to form a data set, preprocess the parameters in the data set, and divide the data into a training set and a test set after annotating them. The specific steps for preprocessing the electrical performance parameters are as follows: Clean the electrical performance parameters to remove noise data and outliers; The electrical performance parameters of different dimensions and ranges are normalized so that they are in the same numerical range.

[0071] S2052: Train the lifespan assessment model using the training set, and test the training effect using the test set to obtain the trained lifespan assessment model.

[0072] In this embodiment, the three branches are trained using their respective losses (MSE, NLL, and BCE), while the remaining branches are frozen during training. Based on the different characteristics of the three branches, the cross entropy between the final Sout and the true label is used as the total loss. Only the top-level parameters of the adaptive weight fusion module and the third branch are updated to prevent catastrophic forgetting. Using unlabeled vehicle data, consistency regularization is applied to the consistency loss of the first and second branch outputs to improve generalization.

[0073] S3: The three branches in the life assessment model process the DC internal resistance, constant voltage stage capacity ratio, and constant current stage capacity ratio in parallel, and the results obtained from the three branches are integrated and decided to obtain the outlier detection result.

[0074] Example 2: A second embodiment of the present invention provides a lithium-ion battery outlier warning system, comprising: The data acquisition module is configured to perform electrical performance tests on the lithium-ion battery, including DC internal resistance at different temperatures and different power levels, capacity percentage in the constant voltage phase, and capacity percentage in the constant current phase; A model selection module is configured to match a corresponding life assessment model according to electrical performance test conditions; The outlier prediction module is configured to use the selected life assessment model to capture the life attenuation law of the lithium battery according to the SOC DC internal resistance, the constant voltage stage capacity ratio, and the constant current stage capacity ratio, and obtain the outlier prediction result of the lithium battery. Among them, the law between the SOC DC internal resistance, the constant voltage stage capacity ratio, the constant current stage capacity ratio and the life attenuation of the lithium battery is learned through the multi-head attention mechanism, and the learned life attenuation law is corrected by using the correction mechanism.

[0075] Example 3: A third embodiment of the present invention provides a computer-readable storage medium storing a computer program. The computer program is suitable for being loaded by a processor and executing the steps of the lithium-ion battery outlier warning method as described in the first embodiment of the present invention.

[0076] Example 4: A fourth embodiment of the present invention provides a computer device, comprising: a processor adapted to execute a computer program; A computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the lithium-ion battery outlier warning method as described in the first embodiment of the present invention are implemented.

[0077] Embodiment 5: A fifth embodiment of the present invention provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of the lithium-ion battery outlier warning method described in the first embodiment of the present invention.

[0078] The steps involved in the above embodiments 2, 3, 4 and 5 correspond to those in the method embodiment 1. For the specific implementation methods, please refer to the relevant description part of the embodiment 1.

[0079] Those skilled in the art will appreciate that the units and algorithmic steps of each example described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technical personnel may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application. In the above embodiments, all or part of the embodiments can be implemented using software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments can be implemented in the form of a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to the embodiments of the present application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted via a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber, digital line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data processing device such as a server or data center that integrates one or more available media. Available media can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)). 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 technical object of a person skilled in the art that can be easily conceived of within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A lithium-ion battery outlier warning method, characterized in that: The following steps are involved: Conduct electrical performance tests on lithium-ion batteries to obtain the DC internal resistance, constant voltage stage capacity ratio, and constant current stage capacity ratio at different temperatures and different power levels; Constructing a lifespan assessment model comprising three branches and training the lifespan assessment model, wherein the three branches include a first branch for local outlier detection, a second branch for residual analysis, and a third branch for deep learning; The three branches in the life assessment model process the DC internal resistance, constant voltage stage capacity ratio, and constant current stage capacity ratio in parallel, and the results obtained from the three branches are integrated and decided to obtain the outlier detection results.

2. The lithium-ion battery outlier warning method according to claim 1, wherein: The first branch performs local outlier detection by constructing a matrix diagram. The second branch performs residual analysis by constructing a fitting formula based on the linear attenuation law and sinusoidal fluctuations. The third branch captures the life attenuation law of the lithium battery according to the SOC DC internal resistance, the capacity ratio in the constant voltage stage, and the capacity ratio in the constant current stage to obtain the outlier prediction results of the lithium battery.

3. The lithium-ion battery outlier warning method according to claim 2, wherein: The third branch uses a multi-head attention mechanism to learn the relationship between SOC DC internal resistance, constant voltage stage capacity ratio, constant current stage capacity ratio and lithium battery life attenuation, and adopts a correction mechanism to correct the learned life attenuation law.

4. The lithium-ion battery outlier warning method according to claim 1, wherein: The specific steps for testing the electrical performance of lithium-ion batteries are: Set different current parameters, voltage parameters and temperature parameters; Carry out electrical performance test experiments on charging and discharging of the lithium-ion battery to be tested; The electrical performance test experiment was repeated by changing the current parameters, voltage parameters and temperature parameters to obtain different test results.

5. The lithium-ion battery outlier warning method according to claim 1, wherein: The training steps of the lifespan assessment model are: Obtain multiple sets of lithium-ion battery electrical performance test data to form a data set, preprocess the parameters in the data set, and divide the data into training set and test set after annotation; The lifespan assessment model is trained using the training set, and the training effect is tested using the test set to obtain the trained lifespan assessment model.

6. The lithium-ion battery outlier warning method according to claim 5, characterized in that: The specific steps for preprocessing electrical performance parameters are: Clean the electrical performance parameters to remove noise data and outliers; The electrical performance parameters of different dimensions and ranges are normalized so that they are in the same numerical range.

7. A lithium-ion battery outlier warning system, characterized in that: include: The data acquisition module is configured to perform electrical performance tests on the lithium-ion battery, including DC internal resistance at different temperatures and different power levels, capacity percentage in the constant voltage phase, and capacity percentage in the constant current phase; a model building module configured to build a life assessment model comprising three branches and train the life assessment model, wherein the three branches include a first branch for local outlier detection, a second branch for residual analysis, and a third branch for deep learning; The outlier prediction module is configured as the three branches in the life assessment model to process the DC internal resistance, constant voltage stage capacity ratio, and constant current stage capacity ratio in parallel, and integrate the results obtained from the three branches to make decisions and obtain the outlier detection results.

8. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the lithium-ion battery outlier warning method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor and executing the lithium-ion battery outlier early warning method according to any one of claims 1 to 6.

10. A computer device, characterized in that: include: a processor adapted to execute a computer program; A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the lithium-ion battery outlier warning method according to any one of claims 1 to 6 is implemented.

Images