Battery Abnormality Detection Method, Device, Storage Medium and Program Product

Abstract

The present application provides a battery abnormality detection method, device, storage medium and program product, which relates to the technical field of battery detection. The method includes: obtaining battery parameters of a battery to be detected during charging and discharging; performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected; if there is a cell module with outlier parameters, determining the cell module with outlier parameters as an abnormal cell module. By performing outlier detection through battery parameters, the embodiments of the present application can, through the judgment of balance, discover abnormal cell modules in the battery to be detected, thereby improving the accuracy of battery abnormality detection.

Description

Technical Field

[0001] The present application relates to the technical field of battery detection, and more particularly, to a method, device, storage medium, and program product for detecting battery anomalies. Background Art

[0002] Currently, threshold judgment methods are mostly used for battery anomaly detection. For example, through the battery safety monitoring section (Energy Storage Management System, ESMS) in the Battery Management System (BMS). By monitoring and recording the charging and discharging currents and voltages of battery cells, battery modules, battery clusters, and / or battery stacks. When the above data collected exceeds a pre-set threshold, the system will handle the anomaly by means of alarm or shutting down the battery charging and discharging function. This detection method has a large error. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a method, device, storage medium, and program product for detecting battery anomalies, so as to improve the accuracy of battery anomaly detection.

[0004] In a first aspect, the embodiments of the present application provide a method for detecting battery anomalies, including:

[0005] Obtaining battery parameters of a battery to be detected during charging and discharging;

[0006] Performing outlier detection based on the battery parameters to determine whether there are cell modules with outlier parameters in the battery to be detected;

[0007] If there are cell modules with outlier parameters, determining the cell modules with outlier parameters as abnormal cell modules;

[0008] Wherein, the battery parameters include cell voltage and / or cell temperature; performing outlier detection based on the battery parameters to determine whether there are cell modules with outlier parameters in the battery to be detected includes:

[0009] Obtaining first battery parameters of each cell in the battery to be detected at the end of charging, and obtaining second battery parameters of each cell in the battery to be detected at the end of discharging;

[0010] For each cell, obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging; obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging;

[0011] Determining whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result;

[0012] Obtain a first outlier result based on the first battery parameters corresponding to the battery cell and the first average battery parameters of the battery to be detected at the end of charging, including:

[0013] Use an outlier detection function to perform outlier analysis on the first battery parameters and the first average battery parameters. If it is determined that the battery cell parameters are outliers and the first battery parameters of the battery cell are greater than the first average battery parameters, then determine that the first outlier result of the battery is a high charging outlier; if it is determined that the battery cell parameters are outliers and the first battery parameters of the battery cell are less than the first average battery parameters, then determine that the first outlier result of the battery is a low charging outlier. [[ID=:5]]

[0014] Obtain a second outlier result based on the second battery parameters corresponding to the battery cell and the second average battery parameters of the battery to be detected at the end of discharging, including:

[0015] Use an outlier detection function to perform outlier analysis on the second battery parameters and the second average battery parameters. If it is determined that the battery cell parameters are outliers and the second battery parameters of the battery cell are less than the second average battery parameters, then determine that the first outlier result of the battery is a low discharging outlier; if it is determined that the battery cell parameters are outliers and the second battery parameters of the battery cell are greater than the second average battery parameters, then determine that the first outlier result of the battery is a high discharging outlier.

[0016] In the embodiments of the present application, outlier detection is performed through the first battery parameters at the end of charging and the second battery parameters at the end of discharging. Since the battery parameters at the end of charging and the end of discharging can reflect the battery performance and health status during the charging process and the discharging process, therefore, outlier detection through the battery parameters at the end of charging and the discharging module can improve the accuracy and reliability of anomaly detection, and help to timely discover and handle potential battery failures.

[0017] In any embodiment, determining whether a battery cell is a battery cell with outlier parameters according to the first outlier result and the second outlier result includes:

[0018] If the first outlier result is a high charging outlier and the second outlier result is a low discharging outlier, then determine that the battery is a battery cell with outlier parameters.

[0019] In the embodiments of the present application, if the voltage and / or battery temperature of a certain battery cell is significantly higher than that of other battery cells at the end of charging, and the voltage and / or battery temperature is significantly lower than that of other battery cells at the end of discharging, it indicates that the battery cell is abnormal. Therefore, abnormal battery cells can be more accurately discovered through this feature.

[0020] In any embodiment, determining whether a battery cell is a battery cell with outlier parameters according to the first outlier result and the second outlier result includes:

[0021] If the first outlier result is a low charging outlier and the second outlier result is a low discharging outlier, then determine that the battery is a battery cell with outlier parameters.

[0022] In the embodiments of the present application, if the voltage and / or battery temperature of a certain battery cell is significantly lower than those of other battery cells at the end of charging, and the voltage and / or battery temperature is also significantly lower than those of other battery cells at the end of discharging, it is determined that the battery cell may have abnormalities such as insufficient power, and the battery cell is an abnormal battery cell.

[0023] In any embodiment, determining whether a battery cell is a parameter-outlier battery cell according to the first outlier result and the second outlier result includes:

[0024] If the first outlier result is a charge-low outlier or a charge-high outlier, and / or the second outlier result is a discharge-low outlier or a discharge-high outlier, it is determined that the battery is a parameter-outlier battery cell.

[0025] In the embodiments of the present application, an outlier battery cell can be determined through the first outlier result and / or the second outlier result, and the battery cell is determined as an abnormal battery cell.

[0026] In any embodiment, the battery parameters include the battery module voltage and / or the battery module temperature; performing outlier detection based on the battery parameters to determine whether there is a parameter-outlier battery cell module in the battery to be detected, including:

[0027] Obtaining the third battery parameters of each battery module in the battery to be detected at the end of charging, and obtaining the fourth battery parameters of each battery module in the battery to be detected at the end of discharging;

[0028] For each battery module, obtaining a third outlier result according to the third battery parameter corresponding to the battery module and the third average battery parameter of the battery to be detected at the end of charging; obtaining a fourth outlier result according to the fourth battery parameter at the end of discharging corresponding to the battery module and the fourth average battery parameter of the battery to be detected at the end of discharging;

[0029] If at least one of the third outlier result and the fourth outlier result indicates that the battery module is an outlier, it is determined that there is a parameter-outlier battery cell module in the battery to be detected.

[0030] In the present application, by performing outlier detection on the voltage and / or the battery module temperature of the battery module, when there is a large difference in the battery parameters between a battery module and other battery modules, it is determined that the battery module is an abnormal battery cell module, and the outlier detection can quickly and accurately identify the abnormal battery module.

[0031] In any embodiment, the battery parameters include the battery cell voltage and / or the battery cell temperature; after obtaining the battery parameters of the battery to be detected during charge and discharge, the method further includes:

[0032] Extracting the battery parameters between adjacent battery cells in the battery to be detected from the obtained battery parameters;

[0033] Calculate the first difference degree of battery parameters between adjacent battery cells;

[0034] If the first difference degree is greater than the first preset difference threshold, it is determined that the battery to be detected is abnormal.

[0035] In the embodiments of the present application, under normal circumstances, the battery parameters of adjacent battery cells do not differ much. If the difference is large, it indicates that the battery is abnormal. Therefore, the difference degree of the voltage and / or battery temperature between adjacent battery cells can be calculated to reasonably and accurately determine whether the battery is abnormal.

[0036] In any embodiment, the battery parameters include the battery stack voltage and the battery stack current; after obtaining the battery parameters of the battery to be detected during the charging and discharging process, the method further includes:

[0037] If the battery stack current is less than the preset current threshold, and the battery stack voltage shows an increasing trend, and the change rate of the battery stack voltage is less than the preset change rate threshold, and the duration is greater than the preset duration, it is determined that the battery stack is abnormal.

[0038] The embodiments of the present application can more accurately check whether the battery is abnormal by monitoring the battery stack current and the battery stack voltage and combining the judgment of the duration.

[0039] In any embodiment, the battery parameters include the state of charge of the battery cluster; after obtaining the battery parameters of the battery to be detected during the charging and discharging process, the method further includes:

[0040] Perform difference degree analysis based on the state of charge of multiple battery clusters in the battery stack to obtain the second difference degree;

[0041] If the second difference degree is greater than the second preset difference degree threshold, it is determined that the battery cluster is abnormal.

[0042] The embodiments of the present application can quantify the SOC difference between battery clusters by analyzing the difference degree of the state of charge of multiple battery clusters in the battery stack. When the SOC of a certain battery cluster is significantly different from that of other battery clusters, it may mean that there is an internal fault or performance degradation in the battery cluster. This difference degree-based judgment method is more accurate than the single-threshold judgment because the abnormality of the battery cluster may be manifested as a significant deviation of the SOC, rather than just an abnormality in the absolute value.

[0043] In any embodiment, after determining the battery cell module with abnormal parameters as the abnormal battery cell module, the method further includes:

[0044] Input the battery parameters into a pre-trained abnormal analysis model to obtain the cause of the abnormality output by the abnormal analysis model; wherein, the abnormal analysis model is generated based on a machine learning model, combining the SHAP theory and the information gain IG evaluation method.

[0045] In the embodiments of the present application, by using a machine learning model for abnormal cause information, a large amount of battery parameter data can be utilized for training and learning to discover the potential relationship between data features in the data and battery abnormalities, thereby achieving more accurate abnormal identification. In the scenario of battery abnormality analysis, the SHAP method can help identify which battery parameters have the greatest impact on abnormal judgment, thereby providing valuable insights and enhancing the transparency and credibility of the model.

[0046] In a second aspect, the embodiments of the present application provide a battery abnormality detection device, including:

[0047] A parameter acquisition module, configured to acquire battery parameters of a battery to be detected during charge and discharge;

[0048] A detection module, configured to perform outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected;

[0049] An abnormality determination module, configured to, if there is a cell module with outlier parameters, determine the cell module with outlier parameters as an abnormal cell module;

[0050] The battery parameters include cell voltage and / or cell temperature; specifically, the detection module is configured to:

[0051] Acquire first battery parameters of each cell in the battery to be detected at the end of charging, and acquire second battery parameters of each cell in the battery to be detected at the end of discharging;

[0052] For each cell, obtain a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging; obtain a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging;

[0053] Determine whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result;

[0054] Obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging includes:

[0055] Performing outlier analysis on the first battery parameter and the first average battery parameter by using an outlier detection function. If it is determined that the cell parameter is an outlier and the first battery parameter of the cell is greater than the first average battery parameter, determine that the first outlier result of the battery is a charging-high outlier; if it is determined that the cell parameter is an outlier and the first battery parameter of the cell is less than the first average battery parameter, determine that the first outlier result of the battery is a charging-low outlier;

[0056] Obtaining a second outlier result based on the second battery parameter corresponding to the battery cell and the second average battery parameter at the end of the discharge of the battery to be detected, including:

[0057] Performing outlier analysis on the second battery parameter and the second average battery parameter by using an outlier detection function. If it is determined that the battery cell parameter is an outlier and the second battery parameter of the battery cell is less than the second average battery parameter, then it is determined that the first outlier result of the battery is a low-discharge outlier; if it is determined that the battery cell parameter is an outlier and the second battery parameter of the battery cell is greater than the second average battery parameter, then it is determined that the first outlier result of the battery is a high-discharge outlier.

[0058] In a third aspect, an embodiment of the present application provides an electronic device, including: a processor, a memory, and a bus. Among them, the processor and the memory complete communication with each other through the bus; the memory stores program instructions executable by the processor, and the processor can execute the method of the first aspect by invoking the program instructions.

[0059] In a fourth aspect, an embodiment of the present application provides a non-transitory computer-readable storage medium, including:

[0060] The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause the computer to execute the method of the first aspect.

[0061] In a fifth aspect, an embodiment of the present application provides a computer program product, including computer program instructions. When the computer program instructions are read and run by a processor, the method of the first aspect is executed.

[0062] Other features and advantages of the present application will be described in the subsequent specification, and, in part, will become apparent from the specification, or can be understood by implementing the embodiments of the present application. The objectives and other advantages of the present application can be achieved and obtained through the structures specifically pointed out in the written specification, claims, and drawings. Description of the Drawings

[0063] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0064] Figure 1 A schematic flowchart of a battery anomaly detection method provided by an embodiment of the present application;

[0065] Figure 2 Another schematic flowchart of a battery anomaly detection method provided by an embodiment of the present application;

[0066] Figure 3 Schematic structural diagram of a battery anomaly detection device provided by an embodiment of the present application;

[0067] Figure 4 Schematic structural diagram of an electronic device entity provided by an embodiment of the present application. Detailed implementation manners

[0068] The embodiments of the technical solutions of the present application will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solutions of the present application more clearly, and therefore are only examples and cannot be used to limit the protection scope of the present application.

[0069] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive inclusion.

[0070] In the description of the embodiments of the present application, technical terms such as "first" and "second" are only used to distinguish different objects and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity, specific order or primary-secondary relationship of the indicated technical features. In the description of the embodiments of the present application, "a plurality of" means more than two, unless otherwise specifically defined.

[0071] Referring to "embodiment" herein means that the specific features, structures or characteristics described in connection with the embodiment may be included in at least one embodiment of this application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.

[0072] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0073] In the description of the embodiments of the present application, the term "a plurality of" refers to more than two (including two). Similarly, "a plurality of groups" refers to more than two groups (including two groups), and "a plurality of pieces" refers to more than two pieces (including two pieces).

[0074] In the description of the embodiments of the present application, unless otherwise clearly specified and limited, technical terms such as "installation", "connection", "linkage", "fixation" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or integrated; it can also be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components or the interaction relationship between two components. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of the present application can be understood according to specific circumstances.

[0075] Battery anomaly refers to the abnormal state or performance degradation of the battery during charging, discharging, or storage. Abnormal scenarios include but are not limited to the following:

[0076] Battery overheating: The battery temperature is too high during use or charging, possibly exceeding the normal operating temperature range.

[0077] Battery short circuit: A direct contact occurs between the positive and negative electrodes inside the battery, resulting in an excessive current, which may cause a fire or explosion.

[0078] Battery swelling or deformation: The battery case swells or deforms due to increased internal pressure, possibly accompanied by liquid leakage.

[0079] Rapid battery power decline: Under normal usage conditions, the battery power rapidly decreases, far lower than the expected usage time.

[0080] Unable to charge: The battery cannot receive the charging current, or the charging speed is extremely slow and cannot be fully charged.

[0081] Battery aging: After long-term use or multiple charge-discharge cycles, the battery performance gradually degrades and the capacity decreases.

[0082] Battery anomalies may lead to serious safety accidents such as fires and explosions. Through timely anomaly detection, problems can be discovered and solved before the accidents occur, enabling the battery to operate more safely and stably. The existing anomaly detection methods set thresholds for some key parameters of the battery, such as temperature thresholds, voltage thresholds, current thresholds, etc. When it is detected that the parameters of the battery cell exceed the set parameter thresholds, it is determined that the battery cell is abnormal. The detection results obtained in this way are not accurate enough because the threshold devices are affected by factors such as battery aging and temperature fluctuations, which may cause changes in the thresholds, thus increasing the risk of misjudgment.

[0083] To solve the above technical problems, an embodiment of the present application provides a battery anomaly detection method, device, storage medium, and program product. Among them, the main body executing this method can be a battery management system or a battery cloud server, etc. If it is a battery cloud server, the battery cloud server can be communicatively connected to the battery to obtain battery parameters during the charging and discharging processes of the battery.

[0084] Moreover, the battery can be a power battery or an energy storage battery. One structure of the battery can include multiple battery stacks, and the multiple battery stacks are connected in series, parallel, or a combination of series and parallel. Each battery stack includes multiple battery clusters, and the multiple battery clusters are connected in series, parallel, or a combination of series and parallel. Each battery cluster includes multiple battery modules connected in series, parallel, or a combination of series and parallel. Each battery module includes multiple battery cells connected in series, parallel, or a combination of series and parallel. Another structure of the battery can include multiple battery cells connected in series, parallel, or a combination of series and parallel. Yet another structure of the battery can include multiple battery stacks, and each battery stack includes multiple battery cells. The structure of the battery can also be other, and the embodiments of the present application do not make specific limitations on this.

[0085] Figure 1 The flow diagram of a battery anomaly detection method provided by an embodiment of the present application is shown in Figure 1 As shown, this method includes:

[0086] Step 101: Obtain the battery parameters of the battery to be detected during the charging and discharging processes;

[0087] Step 102: Perform an outlier detection based on the battery parameters to determine whether there are cell modules with outlier parameters in the battery to be detected;

[0088] Step 103: If there are cell modules with outlier parameters, determine the cell modules with outlier parameters as abnormal cell modules.

[0089] In a specific implementation process, the battery parameters can be relevant parameters of the battery to be detected collected by the battery management system during the charging process and the discharging process, and these parameters can reflect the performance and health status of the battery. The battery parameters can be parameters such as the voltage, current, and battery temperature of each battery cell included in the battery to be detected, and can also be parameters such as the voltage, current, and battery temperature of the battery module, and can also be parameters such as the voltage, current, and battery temperature of the battery cluster. The battery parameters can also include battery capacity, battery internal resistance, etc. Therefore, which specific parameters the battery parameters include can be set according to the actual situation. It should be noted that before obtaining the battery parameters, a document configuration parsing template for the charging and discharging original data can be exported from the host computer. The document configuration parsing template determines which specific contents the battery parameters include. Then, the charging and discharging original data document within the detection time period is exported from the host computer, and the battery parameters are obtained from the charging and discharging original data document according to the document configuration parsing template.

[0090] When obtaining battery parameters, professional test equipment can be used, such as: battery management systems, data acquisition devices, etc. These devices can accurately measure various parameters of the battery during charging and discharging processes and record them for subsequent analysis.

[0091] Outlier detection is used to identify points in the collected dataset that are significantly different from most data points. After obtaining the battery parameters, the battery parameters can be preprocessed, including steps such as data cleaning, missing value filling, and data standardization. This helps improve the accuracy and consistency of the data and lays a foundation for subsequent analysis.

[0092] After obtaining the battery parameters, outlier detection is performed based on the battery parameters to determine whether there are cell modules with parameter outliers in the battery to be detected. It should be noted that the cell module can be a single cell, a battery module composed of multiple cells, a battery cluster composed of multiple battery modules, etc. A cell module with parameter outliers refers to a cell module whose battery parameters are significantly different from those of other cell modules. Whether a significant difference is formed can be determined by some outlier algorithms. For example: it can be simple statistical analysis, 3 Principles, box plot analysis, or machine learning algorithms, etc., are used to analyze the battery parameters to identify outliers that are significantly different from most data points. These outliers may represent cell modules with abnormal performance.

[0093] After passing the outlier detection, if it is found that there are cell modules with parameter outliers in the battery, then the cell module with parameter outliers is determined as an abnormal cell module.

[0094] It should be noted that if it is determined that there are abnormal cell modules in the battery to be detected, then the abnormal information is displayed for the staff to view and determine whether professional personnel are needed to analyze the abnormal details. If professional personnel are needed for analysis, it is handed over to the professional personnel to output an analysis report.

[0095] In the embodiment of the present application, outlier detection is used for anomaly identification. Compared with the traditional method of using thresholds for anomaly identification, the advantage is that it improves the accuracy of detection. For example: for temperature detection, if a normally operating battery is in a high-temperature environment, during the charging and discharging process of each cell, its temperature is generally relatively high. If a temperature threshold is used, it is very likely that the battery will be considered abnormal. However, using the outlier detection method of the present application, if the temperature of the battery cells is generally high but does not reach the temperature of thermal runaway, then it can be considered normal.

[0096] In the embodiments of the present application, by obtaining battery parameters during the charge and discharge processes of the battery to be detected and performing outlier detection based on these parameters, it is possible to accurately determine whether there are cell modules with outlier parameters in the battery pack and take corresponding processing measures. This is of great significance for ensuring the performance and safety of the battery pack.

[0097] Among them, the battery parameters include cell voltage and / or cell temperature; performing outlier detection based on the battery parameters to determine whether there are cell modules with outlier parameters in the battery to be detected includes:

[0098] Obtaining the first battery parameters of each cell in the battery to be detected at the end of charging, and obtaining the second battery parameters of each cell in the battery to be detected at the end of discharging;

[0099] For each cell, obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging; obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging;

[0100] Determining whether the cell is a cell with outlier parameters according to the first outlier result and the second outlier result.

[0101] In a specific implementation process, the battery parameters include cell voltage and / or cell temperature. Therefore, the battery parameters can include only cell voltage, only cell temperature, or both cell voltage and cell temperature. The following describes the above three cases separately:

[0102] (1) The battery parameters include only cell voltage:

[0103] When obtaining the cell voltage of the battery to be detected, the cell voltage of the battery to be detected at the end of charging (referred to as the first battery parameter in the embodiments of the present application) and the cell voltage of the battery to be detected at the end of discharging (referred to as the second battery parameter in the embodiments of the present application) can be obtained.

[0104] For each cell included in the battery to be detected, outlier detection is performed according to the cell voltage at the end of charging corresponding to the cell and the average cell voltage at the end of charging corresponding to the battery to be detected, and a first outlier result of the cell is obtained. The first outlier result is used to characterize whether the cell voltage of the cell at the end of charging is an outlier.

[0105] Similarly, outlier detection is performed according to the cell voltage at the end of discharging corresponding to the cell and the average cell voltage at the end of discharging corresponding to the battery to be detected, and a second outlier result of the cell is obtained. The second outlier result is used to characterize whether the cell voltage of the cell at the end of discharging is an outlier.

[0106] After obtaining the first outlier result and the second outlier result, it can be determined whether the battery cell is a battery cell with outlier parameters based on the first outlier result and / or the second outlier result.

[0107] It should be noted that the end of charging refers to the stage when the battery is close to full charge during the charging process. Specifically, it can be determined by the SOC value, voltage value, etc. of the battery during the charging process. For example, when the SOC value reaches 95%, it can be considered that the end of charging is reached. It can also be when the voltage value of the battery reaches 3.4V, it can be considered that the end of charging is reached. Among them, the SOC value and voltage value for determining the end of charging can be set according to the actual situation, and the embodiments of the present application do not make specific limitations on this.

[0108] During the charging process of the battery, after reaching the end of charging, the charging will continue until the cut-off charging is reached. Therefore, during the process from the end of charging to the cut-off charging, the battery cell voltages at multiple time points can be collected, and then the average value is taken as the battery cell voltage at the end of charging. Of course, the first battery cell voltage collected after entering the end of charging can also be used as the battery cell voltage at the end of charging.

[0109] (2) The battery parameters only include the battery cell temperature:

[0110] When obtaining the battery cell temperature of the battery to be detected, the battery cell temperature of the battery to be detected at the end of charging (referred to as the first battery parameter in the embodiments of the present application) and the battery cell temperature of the battery to be detected at the end of discharging (referred to as the second battery parameter in the embodiments of the present application) can be obtained.

[0111] For each battery cell included in the battery to be detected, outlier detection is performed based on the battery cell temperature at the end of charging corresponding to the battery cell and the average battery cell temperature at the end of charging corresponding to the battery to be detected, and the first outlier result of the battery cell is obtained. The first outlier result is used to characterize whether the battery cell temperature at the end of charging of the battery cell is an outlier.

[0112] Similarly, outlier detection is performed based on the battery cell temperature at the end of discharging corresponding to the battery cell and the average battery cell temperature at the end of discharging corresponding to the battery to be detected, and the second outlier result of the battery cell is obtained. The second outlier result is used to characterize whether the battery cell temperature at the end of discharging of the battery cell is an outlier.

[0113] After obtaining the first outlier result and the second outlier result, it can be determined whether the battery cell is a battery cell with outlier parameters based on the first outlier result and / or the second outlier result.

[0114] It should be noted that the end of charging refers to the stage when the battery is close to full charge during the charging process. Specifically, it can be determined by the SOC value, voltage value, etc. of the battery during the charging process. For example, when the SOC value reaches 95%, it can be considered that the end of charging is reached. It can also be when the voltage value of the battery reaches 3.4V, then it can be considered that the end of charging is reached. Among them, the SOC value and voltage value for judging the end of charging can be set according to the actual situation, and the embodiments of the present application do not make specific limitations on this.

[0115] During the charging process of the battery, when the end of charging is reached, the charging will continue until the charging is cut off. Therefore, during the process from the end of charging to the cut-off of charging, the cell temperature at multiple time points can be collected, and then the average value can be taken as the cell temperature at the end of charging. Of course, the first cell temperature collected after reaching the end of charging can also be used as the cell temperature at the end of charging.

[0116] (3)The battery parameters include cell voltage and cell temperature:

[0117] For the case where the battery parameters include cell voltage and cell temperature, the above method can be used to perform outlier detection on the cell voltage and cell temperature respectively. Thus, the first outlier result and the second outlier result of the cell voltage can be obtained; and the first outlier result and the second outlier result of the cell temperature can be obtained.

[0118] When judging whether a cell is a cell with outlier parameters, it can be determined by comprehensively considering the first outlier result and the second outlier result of the cell voltage, and the first outlier result and the second outlier result of the cell temperature. For example: when it is determined that both the cell voltage and the cell temperature are outliers, it is determined that the cell is an outlier; it can also be that when any one of the cell voltage and the cell temperature is an outlier, it is considered that the cell is an outlier. It can also use the method of weighted average to perform outlier judgment.

[0119] The embodiments of the present application perform outlier detection through the first battery parameters at the end of charging and the second battery parameters at the end of discharging. Since the battery parameters at the end of charging and the end of discharging can reflect the battery performance and health status during the charging process and the discharging process, therefore, performing outlier detection through the battery parameters at the end of charging and the discharging module can improve the accuracy and reliability of anomaly detection, and help to timely discover and handle potential battery failures.

[0120] Among them, obtaining the first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging includes:

[0121] Perform outlier analysis on the first battery parameter and the first average battery parameter using an outlier detection function. If it is determined that the cell parameter is an outlier and the first battery parameter of the cell is greater than the first average battery parameter, then determine that the first outlier result of the battery is a high charge outlier; if it is determined that the cell parameter is an outlier and the first battery parameter of the cell is less than the first average battery parameter, then determine that the first outlier result of the battery is a low charge outlier.

[0122] Obtain a second outlier result based on the second battery parameter corresponding to the cell and the second average battery parameter at the end of discharge of the battery to be detected, including:

[0123] Perform outlier analysis on the second battery parameter and the second average battery parameter using an outlier detection function. If it is determined that the cell parameter is an outlier and the second battery parameter of the cell is less than the second average battery parameter, then determine that the first outlier result of the battery is a low discharge outlier; if it is determined that the cell parameter is an outlier and the second battery parameter of the cell is greater than the second average battery parameter, then determine that the first outlier result of the battery is a high discharge outlier.

[0124] In a specific implementation process, for the outlier detection at the end of battery charging, taking the cell voltage as an example, the cell voltage of a cell with an outlier parameter may be significantly higher than the cell voltages of other cells, or may be significantly lower than the cell voltages of other cells. Therefore, the first outlier result includes a high charge outlier and a low charge outlier. A high charge outlier means that for a certain cell at the end of charging, its battery parameter is significantly higher than that of most other cells. A low charge outlier means that for a certain cell at the end of charging, its battery parameter is significantly lower than that of most other cells.

[0125] Correspondingly, for the outlier detection at the end of battery discharge, the second outlier result can include a high discharge outlier and a low discharge outlier. A high discharge outlier means that for a certain cell at the end of discharge, its battery parameter is significantly higher than that of most other cells. A low discharge outlier means that for a certain cell at the end of discharge, its battery parameter is significantly lower than that of most other cells.

[0126] The judgment basis for a high charge outlier is: perform outlier analysis on the first battery parameter and the first average battery parameter using an outlier detection function. If it is determined that the cell parameter is an outlier and the first battery parameter of the cell is greater than the first average battery parameter.

[0127] The judgment basis for a low charge outlier is: perform outlier analysis on the first battery parameter and the first average battery parameter using an outlier detection function. If it is determined that the cell parameter is an outlier and the first battery parameter of the cell is less than the first average battery parameter.

[0128] The judgment basis for a high discharge outlier is: perform outlier analysis on the second battery parameter and the second average battery parameter using an outlier detection function. If it is determined that the cell parameter is an outlier and the second battery parameter of the cell is greater than the second average battery parameter.

[0129] The basis for determining a low discharge outlier is as follows: an outlier analysis is performed on the second battery parameter and the second average battery parameter using an outlier detection function. If it is determined that the cell parameter is an outlier and the second battery parameter of the cell is less than the second average battery parameter.

[0130] Among them, the outlier detection parameter is preset, for example, it can be: . Among them, is the scenario coefficient, and for different anomaly detection scenarios, the scenario coefficient is different, and this value can be preset; is the degree constant, which is also preset according to different anomaly detection scenarios. The larger this value is, the stricter the detection is; is the outlier value; is the battery parameter of the i-th cell; is the average parameter of the cell (which can be the first average battery parameter, the second average battery parameter, etc.). The outlier value can be calculated through the above formula, and whether the cell is an outlier can be further determined according to the outlier value.

[0131] In addition, the outlier detection parameter can also be based on the 3 principle. The standard deviation is calculated according to the first battery parameter and the first average battery parameter μ, and then the standard parameter interval is determined according to the first average battery parameter and the standard deviation , that is, (μ - 3 , μ + 3 ). If the first battery parameter is within this standard parameter interval, it means it is not an outlier, otherwise it means the battery parameter of this cell is an outlier.

[0132] It should be noted that other outlier detection methods can also be used, and the embodiments of the present application do not make specific limitations in this regard.

[0133] The embodiments of the present application provide a basis for subsequent abnormal detection of the cell module through the determination of overcharge high outlier, overcharge low outlier, discharge high outlier, and discharge low outlier.

[0134] Based on the above embodiments, determining whether a cell is a cell with an outlier parameter according to the first outlier result and the second outlier result includes:

[0135] If the first outlier result is overcharge high outlier and the second outlier result is discharge low outlier, it is determined that the battery is a cell with an outlier parameter.

[0136] In a specific implementation process, if the battery voltage and / or battery temperature of a certain battery cell is higher than that of most other battery cells at the end of charging, and the battery voltage and / or battery temperature of the battery cell is lower than that of most other battery cells at the end of discharging, it may be due to the low capacity of the battery cell, or the battery cells are of mixed grades, or the battery cell is aged. Therefore, the battery cell is determined as a battery cell with abnormal parameters.

[0137] In the embodiment of the present application, if the voltage and / or battery temperature of a certain battery cell is significantly higher than that of other battery cells at the end of charging, and the voltage and / or battery temperature of the battery cell is significantly lower than that of other battery cells at the end of discharging, it indicates that the battery cell is abnormal. Therefore, abnormal battery cells can be more accurately detected through this feature.

[0138] Based on the above embodiments, determining whether a battery cell is a battery cell with abnormal parameters according to the first outlier result and the second outlier result includes:

[0139] If the first outlier result is a charging low outlier and the second outlier result is a discharging low outlier, determine that the battery is a battery cell with abnormal parameters.

[0140] In a specific implementation process, if the battery voltage and / or battery temperature of a certain battery cell is lower than that of most other battery cells at the end of charging, and the battery voltage and / or battery temperature of the battery cell is lower than that of most other battery cells at the end of discharging, it may be due to insufficient power of the battery cell (excessive self-discharge / internal micro-short circuit), or the battery cells are of mixed grades, or the battery cell is aged. Therefore, the battery cell is determined as a battery cell with abnormal parameters. Therefore, the battery cell is determined as a battery cell with abnormal parameters.

[0141] In the embodiment of the present application, if the voltage and / or battery temperature of a certain battery cell is significantly lower than that of other battery cells at the end of charging, and the voltage and / or battery temperature of the battery cell is also significantly lower than that of other battery cells at the end of discharging, it is determined that the battery cell may have abnormalities such as insufficient power, and the battery cell is an abnormal battery cell.

[0142] Based on the above embodiments, determining whether a battery cell is a battery cell with abnormal parameters according to the first outlier result and the second outlier result includes:

[0143] If the first outlier result is a charging low outlier or a charging high outlier, and / or the second outlier result is a discharging low outlier or a discharging high outlier, determine that the battery is a battery cell with abnormal parameters.

[0144] In a specific implementation process, for single battery cell outlier determination, if at least one of a charging high outlier, a charging low outlier, a discharging high outlier, and a discharging low outlier occurs, it can be determined as a single battery cell outlier. Among them, when judging outliers, the outlier judgment rule is relatively strict. For example: using When making a judgment, k can take 1.25 and a can take 1.5. Of course, other values can also be used.

[0145] In the embodiment of the present application, an outlier cell can be determined through the first outlier result and / or the second outlier result, and this cell is determined as an abnormal cell.

[0146] Based on the above embodiment, the battery parameters include the battery module voltage and / or the battery module temperature; outlier detection is performed based on the battery parameters to determine whether there is a cell module with parameter outliers in the battery to be detected, including:

[0147] Obtain the third battery parameter of each battery module in the battery to be detected at the end of charging, and obtain the fourth battery parameter of each battery module in the battery to be detected at the end of discharging;

[0148] For each battery module, obtain a third outlier result according to the third battery parameter corresponding to the battery module and the third average battery parameter of the battery to be detected at the end of charging; obtain a fourth outlier result according to the fourth battery parameter at the end of discharging corresponding to the battery module and the fourth average battery parameter of the battery to be detected at the end of discharging;

[0149] If at least one of the third outlier result and the fourth outlier result indicates that the battery module is an outlier, it is determined that there is a cell module with parameter outliers in the battery to be detected.

[0150] In the specific implementation process, for the abnormal detection of the battery modules in the battery cluster, it is mainly a horizontal comparison between battery modules. Therefore, the battery parameters include the battery module voltage and / or the battery module temperature. A battery module includes multiple cells, and the battery module voltage is the average value of the voltages of all cells in the battery module. The battery module temperature is the average value of the temperatures of all cells in the battery module.

[0151] For a certain battery module, obtain the third battery parameter (battery module voltage and / or battery module temperature) of this battery module at the end of charging, and the corresponding third average battery parameter of the battery cluster where this battery module is located at the end of charging. Use the third battery parameter and the third average battery parameter to perform outlier analysis on this battery module to obtain a third outlier result. Among them, the third outlier result can be that the battery module is highly outlier during charging or that the battery module is lowly outlier during charging.

[0152] Similarly, obtain the fourth battery parameter (battery module voltage and / or battery module temperature) of this battery module at the end of discharging, and the corresponding fourth average battery parameter of the battery cluster where this battery module is located at the end of discharging. Use the fourth battery parameter and the fourth average battery parameter to perform outlier analysis on this battery module to obtain a fourth outlier result. Among them, the fourth outlier result can be that the battery module is highly outlier during discharging or that the battery module is lowly outlier during discharging.

[0153] It should be noted that the obtaining methods of the third outlier result and the fourth outlier result can refer to the obtaining methods of the first outlier result and the second outlier result in the above embodiments, and the embodiments of the present application do not make specific limitations thereto.

[0154] During the charging process of the battery, when the end of charging is reached, the charging will continue until the charging is cut off. Therefore, in the process from the end of charging to the cut-off of charging, for each battery module, the battery module voltages at multiple time points can be calculated, and then the average value can be taken as the battery module voltage at the end of charging. Of course, the first battery module voltage obtained after entering the end of charging can also be used as the battery module voltage at the end of charging. The obtaining method of the battery module temperature is similar and will not be elaborated here.

[0155] During the discharging process of the battery, when the end of discharging is reached, the discharging will continue until the discharging is cut off. Therefore, in the process from the end of discharging to the cut-off of discharging, for each battery module, the battery module voltages at multiple time points can be calculated, and then the average value can be taken as the battery module voltage at the end of discharging. Of course, the first battery module voltage obtained after entering the end of discharging can also be used as the battery module voltage at the end of discharging. The obtaining method of the battery module temperature is similar and will not be elaborated here.

[0156] The present application performs outlier detection on the voltage of the battery module and / or the battery module temperature. When there is a large difference in the battery parameters between a battery module and other battery modules, it is determined that the battery module is an abnormal cell module. Through outlier detection, abnormal battery modules can be quickly and accurately identified.

[0157] Based on the above embodiments, the battery parameters include the cell voltage and / or the cell temperature; after obtaining the battery parameters of the battery to be detected during the charge and discharge processes, the method further includes:

[0158] Extracting the battery parameters between adjacent cells in the battery to be detected from the obtained battery parameters;

[0159] Calculating the first difference degree of the battery parameters between adjacent cells;

[0160] If the first difference degree is greater than the first preset difference threshold, it is determined that the battery to be detected is abnormal.

[0161] In a specific implementation process, the first degree of difference refers to the degree of difference in battery parameters between adjacent battery cells. This degree of difference can be obtained by calculating the absolute difference or relative difference of the battery parameters of adjacent battery cells. Among them, the absolute difference method is to directly calculate the absolute value of the difference between the parameters of adjacent battery cells, that is, the first degree of difference = |battery parameter of cell A - battery parameter of cell B|. The relative difference method refers to the ratio of the difference between two data to their average value, usually expressed in the form of a percentage. The specific formula is: the first degree of difference = , where A is the battery parameter of cell A and B is the battery parameter of cell B.

[0162] The first preset difference threshold is a preset value, and its specific value can be set according to the normal operating range, design rules, and safety standards of the battery to be detected. It is intended to capture significant differences that may indicate abnormalities inside the battery. If the first degree of difference is greater than the first preset difference threshold, it means that the battery parameters (battery voltage / battery temperature) between two adjacent battery cells differ greatly, so it can be determined that the battery to be detected is abnormal.

[0163] In the embodiments of the present application, under normal circumstances, the battery parameters of adjacent battery cells do not differ much. If the difference is large, it means that the battery is abnormal. Therefore, by calculating the degree of difference in voltage and / or battery temperature between adjacent battery cells, it is possible to reasonably and accurately determine whether the battery is abnormal.

[0164] Based on the above embodiments, the battery parameters include the battery stack voltage and the battery stack current; after obtaining the battery parameters of the battery to be detected during charge and discharge, the method further includes:

[0165] If the battery stack current is less than the preset current threshold, and the battery stack voltage shows an increasing trend, and the rate of change of the battery stack voltage is less than the preset rate of change threshold, and the duration is greater than the preset duration, then it is determined that the battery stack is abnormal.

[0166] Based on the above embodiments, a battery stack is usually composed of multiple battery cells, and these battery cells can be of the same type or different types of batteries. In series connection, the voltages of the battery cells are added to obtain the total voltage of the battery stack; in parallel connection, the currents of the battery cells are added, but ideally, the voltage remains unchanged. The battery stack current refers to the total current flowing through the battery stack, and this current is generated by a battery stack composed of multiple battery cells connected in series or parallel during charge and discharge.

[0167] For the abnormal scenario of long-term small-current charging, it can be judged according to the obtained battery stack voltage and battery stack current according to the following conditions:

[0168] (1) Whether the battery stack current is less than the preset current threshold;

[0169] (2) Whether the battery stack voltage shows an increasing trend;

[0170] (3) Whether the battery stack voltage change rate is less than a preset change rate threshold;

[0171] (4) Whether the duration of the phenomena in (1)-(3) above is greater than a preset duration.

[0172] If (1)-(4) are all judged to be yes, it indicates that the battery stack is in an abnormal situation of continuous charging with a small current, and there may be an abnormality in the management logic of the Power Conversion System (PCS).

[0173] The specific values of the preset current threshold, the preset change rate threshold, and the preset duration can be set according to the design situation.

[0174] By monitoring the battery stack current and the battery stack voltage in this embodiment of the application and combining the judgment of the duration, it is possible to more accurately check whether the battery is abnormal.

[0175] Based on the above embodiment, the battery parameters include the state of charge of the battery cluster; after obtaining the battery parameters of the battery to be detected during charging and discharging, the method further includes:

[0176] Performing difference analysis based on the state of charge of multiple battery clusters in the battery stack to obtain a second difference degree;

[0177] If the second difference degree is greater than a second preset difference degree threshold, it is determined that the battery cluster is abnormal.

[0178] In the specific implementation process, the state of charge of the battery cluster refers to the percentage of the current remaining power of the battery cluster in its rated capacity. The state of charge SOC of the battery cluster = (the current power of the battery cluster - the lower limit of the battery capacity) / (the upper limit of the battery capacity - the lower limit of the battery capacity) × 100%.

[0179] After obtaining the state of charge (SOC) of multiple battery clusters within a battery stack, analyze based on the SOC inconsistency among the battery clusters. Specifically, the second degree of difference in the SOC between any two of the multiple battery clusters can be analyzed. Among them, the calculation of the second degree of difference can be obtained by subtracting the SOCs of the two battery clusters. In addition, the calculated second degree of difference can be normalized so as to more intuitively represent the difference length of the SOC between the two battery clusters. The method of normalization is usually to divide the difference value by a reference value (such as the rated capacity of the battery cluster or the SOC value when fully charged) to obtain a degree of difference index between 0 and 1 (or adjust the range according to specific circumstances). However, in this scenario, since the SOC itself is a percentage value and the rated capacities of the two battery clusters may be the same or similar, the difference value can be directly used as the degree of difference, or divided by 100 to obtain a more intuitive percentage form.

[0180] After calculating the second degree of difference, it can be determined whether the second degree of difference is greater than the second preset difference threshold. If it is greater, it indicates that there are battery clusters with a large difference in the state of charge in the battery stack. This may be due to different degrees of aging of the battery cells, resulting in obvious differences between different battery clusters.

[0181] By analyzing the degree of difference in the state of charge of multiple battery clusters within the battery stack in the embodiments of the present application, the SOC difference between each battery cluster can be quantified. When the SOC of a certain battery cluster is significantly different from that of other battery clusters, it may mean that there is an internal fault or performance degradation in this battery cluster. This method of judgment based on the degree of difference is more accurate than the single-threshold judgment because the abnormality of the battery cluster may be manifested as a significant deviation of the SOC, rather than just an abnormality in the absolute value.

[0182] On the basis of the above embodiments, after determining the battery cell module with outlier parameters as an abnormal battery cell module, the method further includes:

[0183] Input the battery parameters into a pre-trained anomaly analysis model to obtain the cause of the anomaly output by the anomaly analysis model; among them, the anomaly analysis model is generated based on a machine learning model, combined with the SHAP theory and the information gain IG evaluation method.

[0184] In the specific implementation process, the embodiments of the present application rely on a machine learning solution. Based on the DecisionTree machine learning model, through machine learning, the manufacturer's outlier function corresponding to the scenario is fitted. The implementation method of this function is to form, through feature engineering, adopt the SHAP theory combined with the information gain IG evaluation method, extract and clean the original data documents marked by the business, summarize the scenario performance characteristics of the same manufacturer and the same cathode material of the battery cell, and extract and screen the feature subset that has the most influence on the battery cell anomaly detection.

[0185] The SHAP (SHapley Additive exPlanations) value is a method based on cooperative game theory for explaining the contributions of individual features in a single prediction value. The SHAP value can provide a more fine-grained assessment of feature importance and is applicable to complex machine learning models. To analyze the causes of battery failures, the embodiments of this application infer the causes of failures through the SHAP theory, providing effective guidance measures for the next step of battery safety maintenance.

[0186] Among them, the calculation formula of the SHAP value is as follows:

[0187]

[0188] represents the k-th subset. Here, the subset means selecting some features from all features, and the selected part of the features is taken as a subset. represents the SHAP value of the i-th feature on the given model f and input instance x. The SHAP value indicates how much the marginal contribution to this specific prediction is when only considering the i-th feature. Σ represents the sum over all possible subsets S. means that S is a subset selected from all feature sets [N] that do not contain feature i. N is the set of all features, and represents the remaining feature set after removing the i-th feature.

[0189] represents the total number of all possible subsets S. Since there are M features and each feature can be selected to be included or not included in the subset, the total number of subsets is . However, since the case containing feature i is excluded here, the actual number of subsets is . represents the number of elements in the k-th subset.

[0190] This part is the weight factor, indicating that among all possible subsets, the probability that the k-th subset of size appears in the set of size M. It is derived from the combination formula in combinatorial mathematics.

[0191] represents the result predicted by the model f when only the subset S and the value of feature i exist. In other words, this is the predicted output corresponding to a set of input vectors that only include the features in S and feature i. Denotes the result predicted by the model f when only the eigenvalues in the subset S exist. That is, this is the predicted output corresponding to an input vector composed only of the features in S. Denotes the change in the predicted result when feature i is added. It reflects the marginal contribution of feature i on the subset S.

[0192] Information Gain (IG) is a method based on information theory, used to measure the reduction in the uncertainty of a dataset after introducing a certain feature. The model f therein can be calculated using information gain, as shown in the following formula:

[0193]

[0194]

[0195] Among them, is the probability that 𝑌 takes the value of and N is the number of possible values of 𝑌.

[0196]

[0197] Among them, is the probability that X takes the value of and M is the number of possible values of X. is the probability of 𝑌 under the condition of

[0198] In the SHAP theory, compares the differences between the model outputs Y with and without feature i. The accuracy of this way of judging differences by distance is limited. Therefore, in the embodiments of the present application, the information gains with and without feature i are considered. Instead of simply using the difference between the two output Y values as the judgment basis, it is judged by taking the difference after entropy calculation (information gain). At the same time, the advantages of the SHAP theory are utilized, that is, considering the contribution of feature i to the whole relative to each subset to evaluate the contribution degree of this feature, and finally completing the evaluation of the importance of all features.

[0199]

[0200] During the model training process, cross-validation technology is adopted to ensure the stability of the model on different datasets. At the same time, pruning technology is used to avoid overfitting.

[0201] Based on the fitting result, the obtained function is used to construct a combination of multiple decision trees (to improve robustness) using the random forest ensemble learning method.

[0202] ​

[0203] For the weight dictionary w of all models, w = {'Model 1': w1, 'Model 2': w2, …, 'Model N': w n}.

[0204] It should be noted that the data used for training the model are battery parameters with known abnormal causes collected historically.

[0205] Through the above process, an anomaly analysis model for anomaly analysis can be obtained, which can improve the accuracy of anomaly analysis in the embodiments of the present application.

[0206] After obtaining the anomaly analysis model, if it is determined through the above embodiments that there is an abnormal cell module in the battery to be detected, the obtained battery parameters are input into the anomaly analysis model, and the anomaly analysis model outputs the abnormal cause of the battery to be detected.

[0207] In the embodiments of the present application, through the machine learning model for anomaly cause information, a large amount of battery parameter data can be used for training and learning, and the potential relationship between the data features in the data and battery anomalies can be mined, so as to achieve more accurate anomaly recognition. In the battery anomaly analysis scenario, the SHAP method can help identify which battery parameters have the greatest impact on anomaly judgment, thereby providing valuable insights and enhancing the transparency and credibility of the model.

[0208] Figure 2 It is a schematic flowchart of another battery anomaly detection method provided by the embodiments of the present application, as Figure 2 shown, the method includes:[[]]

[0209] Step 201: Obtain battery parameters; the battery parameters of the battery to be detected can be obtained through the BMS, or can be obtained through other battery detection devices. Among them, the battery parameters can include the voltage, temperature, etc. of a single cell, or the voltage, temperature, etc. of a battery module, or the state of charge, voltage, temperature, etc. of a battery cluster, or the voltage, current, temperature, etc. of a battery stack.

[0210] Step 202: Obtain the battery parameters at the end of charging and discharging; step 201 can be the battery parameters for the entire time period, and this step can extract the battery parameters at the end of charging and the battery parameters at the end of discharging from step 201.

[0211] Step 203: Data cleaning; clean the obtained battery parameters, for example: remove outliers, fill in missing values, etc.

[0212] Step 204: Outlier analysis of single cell voltages; The outlier analysis of single cell voltages can include outlier analysis of cell voltages at the end of charging and discharging; difference analysis of cell voltages at the end of charging and discharging; and analysis of the charging-high and discharging-low of cell voltages at the end of charging and discharging. The specific analysis methods can be referred to the above embodiments and will not be elaborated here. Among them, each analysis method can obtain a corresponding analysis result, that is, it can be obtained whether there are single cells with outlier parameters in the battery to be detected. Then step 209 is executed.

[0213] Step 205: Outlier analysis of single cell temperatures; The outlier analysis of single cell temperatures can include: outlier analysis of cell temperatures at the end of charging and discharging; difference detection of cell temperatures at the end of charging and discharging; and analysis of the charging-high and discharging-low of cell temperatures at the end of charging and discharging. The specific analysis methods can be referred to the above embodiments and will not be elaborated here. Among them, each analysis method can obtain a corresponding analysis result, that is, it can be obtained whether there are single cells with outlier parameters in the battery to be detected. Then step 209 is executed.

[0214] Step 206: Module outlier analysis; The module outlier analysis can include: analysis of temperature outliers in the module at the end of charging and discharging and analysis of voltage outliers in the module at the end of charging and discharging. The specific analysis methods can be referred to the above embodiments and will not be elaborated here. Among them, each analysis method can obtain a corresponding analysis result, that is, it can be obtained whether there are battery modules with outlier parameters in the battery to be detected. Then step 209 is executed.

[0215] Step 207: Analysis of continuous charging with small current; The analysis method of continuous charging with small current can be referred to the above embodiments and will not be elaborated here. Then step 209 is executed.

[0216] Step 208: Outlier analysis of the SOC of clusters in the battery stack; Judge the difference degree between the SOCs of battery clusters in the battery stack and perform outlier analysis according to the difference degree.

[0217] Step 209: Record the cell components with outlier parameters; Since there cannot be more than one single cell anomaly, or more than one battery module anomaly, or more than one battery cluster anomaly in the battery to be detected, therefore, all abnormal cell modules can be counted.

[0218] Step 210: Record all abnormal cell modules and abnormal scenarios; According to steps 203 - 206, all abnormal cell modules and the scenarios corresponding to the anomalies can be obtained.

[0219] Step 211: Summarize the error messages and display them on the front-end page.

[0220] The embodiments of the present application can identify files in the original data formats of mainstream host computers in the current market (such as xlsx, db, csv, etc.), and through custom import configuration templates, support the system to identify data such as cell voltage and temperature in the original files of different manufacturers for diagnostic work.

[0221] Different from the mainstream solutions in the market that simply judge the thresholds of cells through the BMS system, this solution uses an independent device to analyze by reading the original charge and discharge data of the power station host computer as the data source. The export file formats, data storage locations, and naming of the same fields of each manufacturer are not exactly the same. Through custom configuration of the import template, business personnel can customize the value formats recognized and imported by the system to read the original charge and discharge data of different manufacturers. Since the host computers of the energy management systems (EMS) of mainstream energy storage power station manufacturers in the current market can export the original charge and discharge data, this solution supports the analysis of mainstream power station data in the market, and the value-taking scheme is universal.

[0222] This solution can be integrated into an installation package and used on the windows system. Business personnel can use the integrated software of this solution during factory testing / project implementation / patrol to detect cells at any time and anywhere. The operation steps are simple. For abnormal data, its charge and discharge data can also be visualized for business personnel to analyze. This solution has ease of use.

[0223] Figure 3 It is a schematic structural diagram of a battery anomaly detection device provided by an embodiment of the present application. This device can be a module, program segment, or code on an electronic device. It should be understood that this device corresponds to the above Figure 1 method embodiment and can execute Figure 1 each step involved in the method embodiment. The specific functions of this device can be referred to the description above. To avoid repetition, the detailed description is appropriately omitted here. The device includes: a parameter acquisition module 301, a detection module 302, and an anomaly determination module 303, where:

[0224] The parameter acquisition module 301 is used to acquire battery parameters during the charge and discharge process of the battery to be detected;

[0225] The detection module 302 is used to perform outlier detection based on the battery parameters to determine whether there are cell modules with outlier parameters in the battery to be detected;

[0226] The anomaly determination module 303 is used to, if there are cell modules with outlier parameters, determine the cell modules with outlier parameters as abnormal cell modules;

[0227] The battery parameters include cell voltage and / or cell temperature; the detection module 302 is specifically used for:

[0228] Obtain the first battery parameters of each battery cell in the battery to be detected at the end of charging, and obtain the second battery parameters of each battery cell in the battery to be detected at the end of discharging;

[0229] For each battery cell, obtain a first outlier result according to the first battery parameter corresponding to the battery cell and the first average battery parameter of the battery to be detected at the end of charging; obtain a second outlier result according to the second battery parameter corresponding to the battery cell and the second average battery parameter of the battery to be detected at the end of discharging;

[0230] Determine whether the battery cell is a battery cell with outlier parameters according to the first outlier result and / or the second outlier result;

[0231] The detection module 302 is specifically configured to:

[0232] Perform outlier analysis on the first battery parameter and the first average battery parameter by using an outlier detection function. If it is determined that the battery cell parameter is an outlier and the first battery parameter of the battery cell is greater than the first average battery parameter, then determine that the first outlier result of the battery is a high charge outlier; if it is determined that the battery cell parameter is an outlier and the first battery parameter of the battery cell is less than the first average battery parameter, then determine that the first outlier result of the battery is a low charge outlier;

[0233] The obtaining the second outlier result according to the second battery parameter corresponding to the battery cell and the second average battery parameter of the battery to be detected at the end of discharging includes:

[0234] Perform outlier analysis on the second battery parameter and the second average battery parameter by using the outlier detection function. If it is determined that the battery cell parameter is an outlier and the second battery parameter of the battery cell is less than the second average battery parameter, then determine that the first outlier result of the battery is a low discharge outlier; if it is determined that the battery cell parameter is an outlier and the second battery parameter of the battery cell is greater than the second average battery parameter, then determine that the first outlier result of the battery is a high discharge outlier.

[0235] Based on the above embodiment, the detection module 302 is specifically configured to:

[0236] If the first outlier result is a high charge outlier and the second outlier result is a low discharge outlier, then determine that the battery is a battery cell with outlier parameters.

[0237] Based on the above embodiment, the detection module 302 is specifically configured to:

[0238] If the first outlier result is a low charge outlier and the second outlier result is a low discharge outlier, then determine that the battery is a battery cell with outlier parameters.

[0239] Based on the above embodiments, the detection module 302 is specifically configured to:

[0240] If the first outlier result is a full - low outlier or a full - high outlier, or the second outlier result is a discharge - low outlier or a discharge - high outlier, then determine that the battery is a cell with parameter outliers.

[0241] Based on the above embodiments, the battery parameters include the battery module voltage and / or the battery module temperature; the detection module 302 is specifically configured to:

[0242] Obtain the third battery parameters of each battery module in the charging end of the battery to be detected, and obtain the fourth battery parameters of each battery module in the discharging end of the battery to be detected;

[0243] For each battery module, obtain a third outlier result according to the third battery parameter corresponding to the battery module and the third average battery parameter of the battery to be detected at the charging end; obtain a fourth outlier result according to the fourth battery parameter corresponding to the battery module at the discharging end and the fourth average battery parameter of the battery to be detected at the discharging end;

[0244] If at least one of the third outlier result and the fourth outlier result indicates that the battery module is an outlier, then determine that there is a cell module with parameter outliers in the battery to be detected.

[0245] Based on the above embodiments, the battery parameters include the cell voltage and / or the cell temperature; the detection module 302 is further configured to:

[0246] Extract the battery parameters between adjacent cells in the battery to be detected from the obtained battery parameters;

[0247] Calculate the first difference degree of the battery parameters between adjacent cells;

[0248] If the first difference degree is greater than the first preset difference threshold, then determine that the battery to be detected is abnormal.

[0249] Based on the above embodiments, the battery parameters include the battery stack voltage and the battery stack current; the detection module 302 is further configured to:

[0250] If the battery stack current is less than the preset current threshold, and the battery stack voltage shows an increasing trend, and the change rate of the battery stack voltage is less than the preset change rate threshold, and the duration is greater than the preset duration, then determine that the battery stack is abnormal.

[0251] Based on the above embodiments, the battery parameters include the state of charge of the battery cluster; the detection module 302 is further configured to:

[0252] Perform a difference analysis based on the state of charge of multiple battery clusters in the battery stack to obtain a second difference degree;

[0253] If the second difference degree is greater than the second preset difference degree threshold, determine that the battery cluster is abnormal.

[0254] Based on the above embodiments, the device further includes an abnormality analysis module for:

[0255] Input the battery parameters into a pre-trained abnormality analysis model to obtain the cause of the abnormality output by the abnormality analysis model; wherein, the abnormality analysis model is generated based on a machine learning model, combined with the SHAP theory and the information gain IG evaluation method.

[0256] Figure 4 Schematic diagram of the entity structure of the electronic device provided by the embodiment of the present application, as Figure 4As shown, the electronic device includes: a processor 401, a memory 402, and a bus 403; wherein, the processor 401 and the memory 402 communicate with each other through the bus 403; the processor 401 is configured to call program instructions in the memory 402 to execute the methods provided in the above method embodiments, for example, including: obtaining battery parameters of a battery to be detected during charge and discharge; performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected; if there is a cell module with outlier parameters, determining the cell module with outlier parameters as an abnormal cell module; the battery parameters include cell voltage and / or cell temperature; the performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected includes: obtaining first battery parameters of each cell in the battery to be detected at the end of charging, and obtaining second battery parameters of each cell in the battery to be detected at the end of discharging; for each cell, obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging; obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging; determining whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result; the obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging includes: performing outlier analysis on the first battery parameter and the first average battery parameter by using an outlier detection function, if it is determined that the cell parameters are outlier and the first battery parameter of the cell is greater than the first average battery parameter, determining the first outlier result of the battery as overcharge outlier; if it is determined that the cell parameters are outlier and the first battery parameter of the cell is less than the first average battery parameter, determining the first outlier result of the battery as undercharge outlier; the obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging includes: performing outlier analysis on the second battery parameter and the second average battery parameter by using the outlier detection function, if it is determined that the cell parameters are outlier and the second battery parameter of the cell is less than the second average battery parameter, determining the first outlier result of the battery as underdischarge outlier; if it is determined that the cell parameters are outlier and the second battery parameter of the cell is greater than the second average battery parameter, determining the first outlier result of the battery as over-discharge outlier.

[0257] The processor 401 may be an integrated circuit chip with signal processing capabilities. The aforementioned processor 401 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0258] The memory 402 may include, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), an Electrically Erasable Programmable Read-Only Memory (EEPROM), etc.

[0259] This embodiment discloses a computer program product. The computer program product includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the methods provided in the above method embodiments. For example, it includes: obtaining battery parameters of a battery to be detected during charging and discharging; performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected; if there is a cell module with outlier parameters, determining the cell module with outlier parameters as an abnormal cell module; the battery parameters include cell voltage and / or cell temperature; the performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected includes: obtaining first battery parameters of each cell in the battery to be detected at the end of charging, and obtaining second battery parameters of each cell in the battery to be detected at the end of discharging; for each cell, obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging; obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging; determining whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result; the obtaining the first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging includes: performing outlier analysis on the first battery parameter and the first average battery parameter by using an outlier detection function. If it is determined that the cell parameters are outliers and the first battery parameter of the cell is greater than the first average battery parameter, determining the first outlier result of the battery as overcharging outlier; if it is determined that the cell parameters are outliers and the first battery parameter of the cell is less than the first average battery parameter, determining the first outlier result of the battery as undercharging outlier; the obtaining the second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging includes: performing outlier analysis on the second battery parameter and the second average battery parameter by using the outlier detection function. If it is determined that the cell parameters are outliers and the second battery parameter of the cell is less than the second average battery parameter, determining the first outlier result of the battery as underdischarging outlier; if it is determined that the cell parameters are outliers and the second battery parameter of the cell is greater than the second average battery parameter, determining the first outlier result of the battery as over-discharging outlier.

[0260] This embodiment provides a non-transitory computer-readable storage medium storing computer instructions that cause the computer to execute the methods provided in the above method embodiments, for example, including: obtaining battery parameters of a battery to be detected during charge and discharge; performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected; if there is a cell module with outlier parameters, determining the cell module with outlier parameters as an abnormal cell module; the battery parameters including cell voltage and / or cell temperature; the performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected includes: obtaining first battery parameters of each cell in the battery to be detected at the end of charging, and obtaining second battery parameters of each cell in the battery to be detected at the end of discharging; for each cell, obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging; obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging; determining whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result; the obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging includes: performing outlier analysis on the first battery parameter and the first average battery parameter by using an outlier detection function, if it is determined that the cell parameters are outliers and the first battery parameter of the cell is greater than the first average battery parameter, determining the first outlier result of the battery as overcharge outlier; if it is determined that the cell parameters are outliers and the first battery parameter of the cell is less than the first average battery parameter, determining the first outlier result of the battery as undercharge outlier; the obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging includes: performing outlier analysis on the second battery parameter and the second average battery parameter by using the outlier detection function, if it is determined that the cell parameters are outliers and the second battery parameter of the cell is less than the second average battery parameter, determining the first outlier result of the battery as underdischarge outlier; if it is determined that the cell parameters are outliers and the second battery parameter of the cell is greater than the second average battery parameter, determining the first outlier result of the battery as over-discharge outlier.

[0261] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For another example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling, direct coupling, or communication connection between each other can be through some communication interfaces. The indirect coupling or communication connection of the devices or units can be in electrical, mechanical, or other forms.

[0262] In addition, the units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0263] Furthermore, in each embodiment of the present application, the various functional modules can be integrated together to form an independent part, or each module can exist alone, or two or more modules can be integrated to form an independent part.

[0264] In this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0265] The above description is only for the embodiments of the present application and is not used to limit the protection scope of the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for detecting battery anomalies, characterized in that, Including: Obtaining battery parameters of a battery to be detected during charge and discharge; Performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected; If there is a cell module with outlier parameters, determining the cell module with outlier parameters as an abnormal cell module; The battery parameters include cell voltage and / or cell temperature; the performing outlier detection based on the battery parameters to determine whether there is a cell module with outlier parameters in the battery to be detected includes: Obtaining first battery parameters of each cell in the battery to be detected at the end of charging, and obtaining second battery parameters of each cell in the battery to be detected at the end of discharging; For each cell, obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging; obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging; Determining whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result; The obtaining a first outlier result according to the first battery parameter corresponding to the cell and the first average battery parameter of the battery to be detected at the end of charging includes: Performing outlier analysis on the first battery parameter and the first average battery parameter by using an outlier detection function. If it is determined that the cell parameter is an outlier and the first battery parameter of the cell is greater than the first average battery parameter, determining the first outlier result of the battery as overcharge outlier; if it is determined that the cell parameter is an outlier and the first battery parameter of the cell is less than the first average battery parameter, determining the first outlier result of the battery as undercharge outlier; The obtaining a second outlier result according to the second battery parameter corresponding to the cell and the second average battery parameter of the battery to be detected at the end of discharging includes: Performing outlier analysis on the second battery parameter and the second average battery parameter by using the outlier detection function. If it is determined that the cell parameter is an outlier and the second battery parameter of the cell is less than the second average battery parameter, determining the second outlier result of the battery as underdischarge outlier; if it is determined that the cell parameter is an outlier and the second battery parameter of the cell is greater than the second average battery parameter, determining the second outlier result of the battery as over-discharge outlier; The outlier detection function is as follows: ; Among them, is the scenario coefficient, and for different anomaly detection scenarios, the scenario coefficients are different; is the degree constant, which is also preset according to different anomaly detection scenarios. The larger this value is, the stricter the detection is; is the outlier; is the battery parameter of the i-th battery cell; is the first average battery parameter or the second average battery parameter of the battery cell.

2. The method according to claim 1, characterized in that, The determining whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result includes: If the first outlier result is overcharge outlier and the second outlier result is underdischarge outlier, determining that the cell is a cell with outlier parameters.

3. The method according to claim 1, characterized in that, The determining whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result includes: If the first outlier result is undercharge outlier and the second outlier result is underdischarge outlier, determining that the cell is a cell with outlier parameters.

4. The method according to claim 1, characterized in that The determining whether the cell is a cell with outlier parameters according to the first outlier result and / or the second outlier result includes: If the first outlier result is a full-low outlier or a full-high outlier, or the second outlier result is a discharge-low outlier or a discharge-high outlier, then it is determined that the battery cell is a battery cell with parameter outliers.

5. The method according to claim 1, wherein The battery parameters include the battery module voltage and / or the battery module temperature; the outlier detection based on the battery parameters to determine whether there is a battery cell module with parameter outliers in the battery to be detected includes: Obtaining the third battery parameters of each battery module in the battery to be detected at the end of charging, and obtaining the fourth battery parameters of each battery module in the battery to be detected at the end of discharging; For each battery module, obtaining a third outlier result according to the third battery parameter corresponding to the battery module and the third average battery parameter of the battery to be detected at the end of charging; obtaining a fourth outlier result according to the fourth battery parameter at the end of discharging corresponding to the battery module and the fourth average battery parameter of the battery to be detected at the end of discharging; If at least one of the third outlier result and the fourth outlier result indicates that the battery module is an outlier, it is determined that there is a battery cell module with parameter outliers in the battery to be detected.

6. The method according to claim 1, wherein The battery parameters include the battery cell voltage and / or the battery cell temperature; after obtaining the battery parameters of the battery to be detected during charging and discharging, the method further includes: Extracting the battery parameters between adjacent battery cells in the battery to be detected from the obtained battery parameters; Calculating a first difference degree of the battery parameters between adjacent battery cells; If the first difference degree is greater than a first preset difference threshold, it is determined that the battery to be detected is abnormal.

7. The method according to claim 1, characterized in that, The battery parameters include the battery stack voltage and the battery stack current; after obtaining the battery parameters of the battery to be detected during charging and discharging, the method further includes: If the battery stack current is less than a preset current threshold, and the battery stack voltage shows an increasing trend, and the change rate of the battery stack voltage is less than a preset change rate threshold, and the duration is greater than a preset duration, it is determined that the battery stack is abnormal.

8. The method according to claim 1, wherein The battery parameters include the state of charge of the battery cluster; after obtaining the battery parameters of the battery to be detected during charging and discharging, the method further includes: Performing difference degree analysis according to the states of charge of multiple battery clusters in the battery stack to obtain a second difference degree; If the second difference degree is greater than a second preset difference degree threshold, it is determined that the battery cluster is abnormal.

9. The method according to claim 1, characterized in that, After determining the battery cell module with parameter outliers as an abnormal battery cell module, the method further includes: Inputting the battery parameters into a pre-trained anomaly analysis model to obtain the anomaly cause output by the anomaly analysis model; wherein, the anomaly analysis model is generated based on a machine learning model, combined with the SHAP theory and the information gain IG evaluation method.

10. An electronic device, characterized in that, Includes: A processor, a memory, and a bus, wherein, The processor and the memory complete communication with each other through the bus; The memory stores program instructions that can be executed by the processor, and the processor can execute the method according to any one of claims 1-9 by calling the program instructions.

11. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions, and when the computer instructions are run by a computer, the computer is caused to execute the method according to any one of claims 1-9.

12. A computer program product, characterized in that, It includes computer program instructions, and when the computer program instructions are read and run by a processor, the method according to any one of claims 1-9 is executed.

Images