Battery cell anomaly detection method, apparatus, device, and medium

Abstract

The application relates to the technical field of data processing, can be applied to business scenes such as battery management, financial technology and medical health, and discloses a battery unit abnormality detection method, device, equipment and medium, which comprises the following steps: acquiring static attribute data and running state parameters of a plurality of battery units, generating a homogenized battery unit group through dynamic clustering; performing dynamic benchmark analysis by using internal resistance data, identifying abnormal battery units and generating a preliminary abnormal battery unit set; classifying the abnormal battery units by using a time series analysis model to generate a confirmed abnormal battery unit list; and generating a battery unit maintenance schedule table based on the confirmed list and a preset maintenance knowledge base. By combining dynamic benchmark analysis and a time series analysis model, the application can accurately identify the abnormal state of the battery unit, can discover potential problem batteries in a timely manner, improves the efficiency and accuracy of the battery monitoring system, and guarantees the reliability of the battery at critical moments.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus, device, and storage medium for detecting abnormalities in battery cells. Background Technology

[0002] In the field of monitoring and management of backup lead-acid batteries for data center UPS systems, traditional monitoring methods mainly rely on setting fixed thresholds to trigger alarms. This method identifies the battery's health status by monitoring its internal resistance, voltage, temperature, and other operating conditions, especially changes in internal resistance. However, because the thresholds are fixed and lack dynamic adjustment capabilities, relying solely on this method often fails to identify batteries in the early stages of degradation in a timely manner. This technical challenge is particularly evident in data center applications. In large-scale battery banks, as the number of batteries increases, the efficiency of manual analysis and monitoring decreases significantly, easily leading to the failure to detect early-stage degradation batteries in a timely manner. Especially when the internal resistance just begins to change, without timely and accurate monitoring methods, the gradual degradation trend of the battery often cannot be identified. Alarms are only triggered when the internal resistance suddenly spikes and reaches the threshold, by which time the battery capacity has often decreased significantly or even lost its normal operating capability.

[0003] Similar monitoring and management issues exist in the fintech sector. Financial institutions typically rely on battery packs in their systems to ensure the power supply of critical infrastructure, especially in emergency situations. However, traditional battery monitoring methods still primarily rely on simple threshold alerts, lacking in-depth analysis and dynamic prediction of battery health status. Battery degradation is often gradual; without accurate trend identification, potential battery failures cannot be detected in advance, potentially leading to delays in restoring financial services during critical moments, thereby impacting system stability and business continuity.

[0004] In the healthcare sector, many hospitals and medical devices rely on lead-acid batteries as emergency power. However, traditional battery monitoring methods suffer from the same problems in medical facilities, particularly in hospital life support equipment and data recording systems. Changes in internal resistance directly affect the equipment's operating time and stability. If the monitoring system cannot accurately identify early battery degradation, the equipment may be unable to provide power support at critical moments, impacting the continuity and safety of medical services. Traditional internal resistance threshold alarm systems not only struggle to detect minute changes in internal resistance in a timely manner, but also often reach the alarm threshold when the battery has already entered a degradation phase and cannot be recovered in time.

[0005] In summary, existing technologies generally suffer from the problems of fixed threshold settings and low efficiency of manual analysis, failing to provide accurate and real-time battery health status assessments. Especially in large-scale applications, manual intervention cannot effectively address the early identification of potential battery failures, potentially leading to battery failure at critical moments and impacting system stability and emergency response capabilities. Therefore, a more intelligent and accurate battery monitoring method is urgently needed to address the volatile situations in battery health management and the management needs of large-scale battery packs. Summary of the Invention

[0006] The main objective of this invention is to provide a method, apparatus, device, and storage medium for detecting abnormalities in battery cells. This invention aims to solve the technical problem that traditional battery monitoring methods rely solely on fixed threshold alarms, which cannot accurately identify potential problems in the early stages of battery degradation. This results in the inability to detect battery deterioration in a timely manner, affecting system stability and emergency response capabilities.

[0007] To achieve the above objectives, the present invention provides a method for detecting abnormalities in battery cells, comprising:

[0008] Obtain static attribute data and operating status parameters of multiple battery cells;

[0009] Based on the static attribute data, the multiple battery cells are dynamically clustered to generate homogeneous battery cell groups.

[0010] Within the homogeneous battery cell group, dynamic benchmark analysis is performed based on the internal resistance data in the operating state parameters to initially identify abnormal battery cells and form a preliminary set of abnormal battery cells.

[0011] For each battery cell in the preliminary abnormal battery cell set, a multivariate time series analysis is performed using a time series analysis model to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells.

[0012] Based on the confirmed abnormal battery cell list and the preset maintenance knowledge base, a battery cell maintenance plan is generated.

[0013] Furthermore, to achieve the above objectives, the present invention provides a battery cell anomaly detection device, comprising:

[0014] The data acquisition module is used to acquire static attribute data and operating status parameters of multiple battery cells;

[0015] The clustering analysis module is used to dynamically cluster the multiple battery cells based on the static attribute data to generate homogeneous battery cell groups.

[0016] The benchmark analysis module is used to perform dynamic benchmark analysis based on the internal resistance data in the operating state parameters within the homogeneous battery cell group, to initially identify abnormal battery cells and form a preliminary set of abnormal battery cells.

[0017] The time series analysis module is used to perform multivariate time series analysis on each battery cell in the preliminary abnormal battery cell set using a time series analysis model, so as to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells.

[0018] The maintenance plan generation module is used to generate a battery cell maintenance plan table based on the confirmed abnormal battery cell list and the preset maintenance knowledge base.

[0019] Furthermore, to achieve the above objectives, the present invention also provides a computer device, the computer device including a memory, a processor, and a battery cell anomaly detection program stored in the memory and executable on the processor, wherein when the battery cell anomaly detection program is executed by the processor, it implements the steps of the battery cell anomaly detection method as described above.

[0020] Furthermore, to achieve the above objectives, the present invention also provides a computer-readable storage medium storing a battery cell anomaly detection program, which, when executed by a processor, implements the steps of the battery cell anomaly detection method as described above.

[0021] Beneficial Effects: This invention relates to the field of data processing technology and can be applied to business scenarios such as battery management, fintech, and healthcare. It discloses a method, apparatus, device, and medium for detecting abnormal battery cells, comprising: acquiring static attribute data and operating status parameters of multiple battery cells; generating a homogeneous battery cell group through dynamic clustering; performing dynamic benchmark analysis based on internal resistance data within the homogeneous battery cell group to initially identify abnormal battery cells and generate a preliminary set of abnormal battery cells; performing multivariate time-series analysis on each battery cell in the preliminary set of abnormal battery cells using a time-series analysis model to generate a list of confirmed abnormal battery cells; and generating a battery cell maintenance plan based on the list of confirmed abnormal battery cells and a preset maintenance knowledge base. This invention, by combining dynamic benchmark analysis and a time-series analysis model, accurately identifies abnormal states of battery cells, avoiding misjudgments caused by fixed thresholds in traditional methods. It can promptly detect potentially problematic batteries, improving the efficiency and accuracy of battery monitoring systems and ensuring battery reliability at critical moments. Attached Figure Description

[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0023] Figure 1This is a schematic diagram of an application environment for a battery cell anomaly detection method according to an embodiment of the present invention;

[0024] Figure 2 This is a flowchart illustrating an embodiment of the battery cell anomaly detection method of the present invention;

[0025] Figure 3 This is a schematic diagram of the functional modules of a preferred embodiment of the battery cell anomaly detection device of the present invention;

[0026] Figure 4 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention;

[0027] Figure 5 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0028] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0029] The battery cell anomaly detection method provided in this invention can be applied to, for example... Figure 1 In this application environment, the user terminal communicates with the server via a network. The server can obtain static attribute data and operating status parameters of multiple battery cells from the user terminal, and generate homogeneous battery cell groups through dynamic clustering. Within the homogeneous battery cell groups, dynamic benchmark analysis is performed based on internal resistance data to initially identify abnormal battery cells and generate a preliminary set of abnormal battery cells. For each battery cell in the preliminary set of abnormal battery cells, multivariate time-series analysis is performed using a time-series analysis model to generate a list of confirmed abnormal battery cells. Based on the list of confirmed abnormal battery cells and a preset maintenance knowledge base, a battery cell maintenance plan is generated. This invention, by combining dynamic benchmark analysis and time-series analysis models, accurately identifies abnormal states of battery cells, avoiding misjudgments caused by fixed thresholds in traditional methods. It can promptly detect potential problematic batteries, improve the efficiency and accuracy of the battery monitoring system, and ensure the reliability of the battery at critical moments. The user terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a standalone server or a server cluster composed of multiple servers. The invention will be described in detail below through specific embodiments.

[0030] Please see Figure 2 , Figure 2 This is a flowchart illustrating an embodiment of the battery cell anomaly detection method provided by the present invention. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0031] like Figure 2 As shown, the battery cell anomaly detection method proposed in this invention includes the following steps:

[0032] S10, acquire static attribute data and operating status parameters of multiple battery cells;

[0033] In this embodiment, when acquiring static attribute data and operational status parameters of multiple battery cells, the static attribute data and operational status parameters play different roles. Static attribute data is the battery's fundamental information, providing insights into its basic configuration and usage history. This data is acquired by the battery management system during the production or installation phase, typically by reading battery tags or querying a database. Operational status parameters, on the other hand, are data collected during actual operation, usually through sensors and monitoring systems in real time. Internal resistance, voltage, temperature, and charge / discharge status are important indicators of battery health, especially changes in internal resistance, which are closely related to battery aging and capacity degradation. These parameters for each battery cell can be obtained using existing sensor devices and real-time monitoring systems. The battery management system periodically collects and stores this data to ensure its usability in subsequent analyses.

[0034] After acquiring the data, standardization and data cleaning can unify data from different sources, making them more consistent and comparable in subsequent analyses. For example, battery internal resistance data may vary at different measurement times and with different measuring devices. Therefore, these data need to be standardized to eliminate the influence of different devices and measurement conditions. Standardization typically employs methods such as z-score standardization or min-max standardization to ensure fair comparison of data between different battery cells.

[0035] In practical applications, acquiring this data typically requires the use of existing battery management systems (BMS) and monitoring equipment. BMS usually integrates sensors that can collect real-time operating status parameters of the battery cells. The battery management system is also responsible for preliminary processing of this data, including data cleaning and preprocessing, to ensure data accuracy and consistency.

[0036] During implementation, the first step is to equip each battery cell with appropriate sensors to collect real-time operating status parameters. These sensors typically include internal resistance sensors, voltage sensors, temperature sensors, and charge / discharge status sensors. Data from these sensors can be transmitted to the battery management system (BMS) for storage and processing via wired or wireless means. Static attribute data is usually obtained by reading battery tags or retrieving it from a database. All data acquisition processes must comply with the standards and specifications of the BMS to ensure data accuracy and real-time performance.

[0037] After data acquisition, the raw data typically needs to be cleaned. Data cleaning includes steps such as removing outliers and filling in missing values ​​to ensure the quality of the data used in subsequent analysis. This process can be achieved by setting reasonable threshold ranges to filter data and remove data points that do not conform to reality. For example, voltage and internal resistance values ​​may contain outliers due to equipment failure or transmission errors, and these must be cleaned at this time.

[0038] After data cleaning, the next step is to standardize the operating parameters. This step ensures data consistency during analysis, especially when parameters from different battery cells differ in scale. Standardization eliminates these differences, making the data from different battery cells comparable. Standardization typically employs methods such as z-score standardization or min-max standardization to ensure that different data ranges are converted to the same scale.

[0039] Example Explanation: In a battery management system (BMS), acquiring static attribute data and operational status parameters of multiple battery cells is fundamental for battery health monitoring and maintenance. For instance, in a data center, the operational status parameters of each UPS backup lead-acid battery include internal resistance, voltage, temperature, and charge / discharge status. This data is collected in real time by the BMS to determine the battery's health status. By acquiring these parameters and combining them with static attribute data, such as the battery's brand, model, capacity, and service life, maintenance personnel can conduct detailed analysis of battery performance. In the early stages of battery performance degradation, by monitoring the changing trend of internal resistance in real time, potential battery problems can be identified early, allowing for maintenance or replacement, preventing problems from being discovered only after battery failure, and ensuring that the data center's UPS system can reliably provide emergency power during power outages.

[0040] In battery management systems for medical devices, the battery performance of each device is crucial for stable operation. By acquiring static attribute data (such as battery model and capacity) and operational parameters (such as battery voltage, temperature, and internal resistance) of battery cells, medical device maintenance teams can monitor battery health in real time and identify potential aging issues. For example, for batteries used in pacemakers, monitoring parameters such as changes in internal resistance and temperature fluctuations can promptly detect battery anomalies, preventing the device from failing to provide power at critical moments. This extends the device's lifespan and allows for proactive replacement planning, thereby improving patient safety and treatment outcomes.

[0041] In financial data centers, UPS batteries are critical components ensuring uninterrupted system operation. By acquiring static attribute data and operational status parameters of battery cells, financial institutions can implement battery health monitoring to ensure sufficient power delivery during system outages, preventing data center system shutdowns due to battery failure. For example, real-time monitoring of battery internal resistance, temperature, voltage, and other data allows for analysis of battery health trends, early identification of potential problems, and adjustment of maintenance strategies based on data change patterns. Financial institutions can then develop more precise battery maintenance plans based on this monitoring data, reducing the risk of battery failure, ensuring the continuous availability of financial services, and preventing service quality impacts from unforeseen events caused by battery failure.

[0042] This embodiment acquires and cleans the static attribute data and operating status parameters of multiple battery cells, laying the foundation for subsequent dynamic clustering analysis and anomaly detection. Through proper data preprocessing, errors and anomalies during the acquisition process can be eliminated, ensuring data quality and reliability. This processing method not only improves the accuracy of subsequent clustering analysis and anomaly detection but also reduces the possibility of human intervention and erroneous judgments, thereby significantly improving the efficiency of battery monitoring and maintenance.

[0043] S20, dynamically cluster the multiple battery cells according to the static attribute data to generate a homogeneous battery cell group;

[0044] In this embodiment, the operation of dynamically clustering multiple battery cells to generate homogeneous battery groups aims to group them based on the static attribute data of the battery cells, so as to enable more accurate performance analysis, fault prediction, and maintenance decisions in the future. Static attribute data refers to parameters that do not change over time, such as the battery's brand, model, capacity specifications, and service life. These attributes provide basic information about the battery itself.

[0045] This operation first requires preprocessing the static attribute data. This includes encoding and standardizing the static attributes of each battery cell, such as brand, model, capacity, and service life. Encoding and standardization eliminate differences between different battery models and brands, allowing subsequent clustering operations to be performed on a uniform scale.

[0046] Next, clustering is performed based on the processed static attribute data. The purpose of clustering is to group battery cells into multiple groups based on similarity, ensuring that battery cells within the same group are highly similar in certain dimensions. This allows for unified monitoring, maintenance, and fault diagnosis of these battery cells. Common clustering methods include K-means clustering and DBSCAN. Here, we use a distance metric to cluster the battery cells. By optimizing the grouping of static attribute data based on distance, we can ensure that the performance and feature similarity of battery cells within homogeneous groups are maximized.

[0047] Following clustering, further verification is required to ensure that the batteries within each battery cell group are functionally identical. This step is crucial for guaranteeing the accuracy of subsequent health monitoring and maintenance plans based on these battery cell groups. By performing attribute consistency verification, it is ensured that the battery cells within homogeneous battery cell groups do not exhibit significant performance deviations, thereby improving the accuracy of analysis and fault diagnosis.

[0048] In practice, the preprocessing of static attribute data includes encoding all categorical data (such as brand and model) and converting them into numerical data. Numerical data (such as capacity and service life) then requires standardization. Standardization methods (such as Z-score standardization or Min-Max standardization) can be used to transform the data to a uniform scale, avoiding the impact of large differences in numerical ranges on clustering results.

[0049] The choice of clustering method can be adjusted according to the actual situation. For small datasets, the classic K-means clustering method can be chosen, with an appropriate number of clusters (K value). For larger datasets, or when it is desired to discover cluster structures of arbitrary shapes, the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm can be used. It does not require pre-setting the number of clusters and can better handle noisy data.

[0050] After the clustering results are obtained, attribute consistency verification is required. For example, by calculating the average values ​​of parameters such as internal resistance, temperature, and voltage of the batteries within each battery cell group, the attribute differences between the batteries in the group can be checked. If there are significant differences in the attributes of the batteries within a certain group, then that group may not be suitable as a homogeneous group and needs to be adjusted or re-divided.

[0051] Example Explanation: In a data center battery management system, the static attributes of each UPS battery unit, such as brand, model, capacity, and service life, directly affect the battery's health and performance. When battery units are clustered using static attribute data, the data center can group batteries with similar characteristics together for unified monitoring and maintenance. If a battery pack exhibits abnormal internal resistance, after clustering, the system can immediately determine the health status of all similar batteries in that group, enabling rapid maintenance or replacement, improving overall operational efficiency, and preventing sudden downtime events.

[0052] In battery management for healthcare devices, clustering battery cells using static attribute data can help identify potential risks of battery failure in advance. This allows for monitoring of the battery during normal device operation, ensuring it operates at its optimal condition and preventing device malfunctions or interruptions to healthcare services due to battery failure.

[0053] In the financial sector, by clustering battery static attributes, battery cells can be managed more effectively across multiple key data points. For example, battery capacity and service life are major factors affecting battery performance. By grouping and managing these battery cells, financial data centers can better monitor battery health and ensure the reliability and stability of uninterruptible power supply systems.

[0054] This embodiment utilizes dynamic clustering of static attribute data to group battery cells with similar performance into the same group, thereby improving the efficiency and accuracy of subsequent analysis. In application scenarios such as data centers, healthcare, and finance, clustering operations enable more precise battery health monitoring and fault diagnosis, thereby optimizing battery maintenance plans, reducing system failure risks, and ensuring the continuous and reliable operation of critical equipment.

[0055] S30, within the homogeneous battery cell group, dynamic benchmark analysis is performed based on the internal resistance data in the operating state parameters to initially identify abnormal battery cells and form a preliminary abnormal battery cell set.

[0056] In this embodiment, within a homogeneous battery cell group, dynamic benchmark analysis is performed based on internal resistance data from operating status parameters. The aim is to identify abnormal battery cells and form a preliminary set of abnormal battery cells based on changes in internal resistance. Internal resistance is a crucial parameter for assessing battery health; an increase in internal resistance directly affects the battery's charge / discharge efficiency and lifespan. Therefore, dynamic benchmark analysis is a key step in battery health management.

[0057] First, before conducting dynamic benchmark analysis, it is necessary to obtain real-time performance information of the battery cells using internal resistance data from the operating status parameters. Internal resistance data is an important indicator reflecting the battery's health status; therefore, it is necessary to collect the battery's internal resistance value in real time using a high-frequency monitoring system. The collected data can be used to generate an internal resistance variation curve for comparative analysis and benchmark construction.

[0058] The core of dynamic benchmark analysis lies in generating a dynamic benchmark curve based on these real-time acquired internal resistance data. The benchmark curve represents the ideal internal resistance range derived from statistical analysis of the battery cell's internal resistance data over a period of time. By comparing and analyzing the benchmark curves, it's possible to determine whether the battery's internal resistance changes are abnormal and further identify battery cells whose internal resistance values ​​deviate from the normal range. This analysis process can be performed across multiple battery cells, ensuring that each battery cell is appropriately positioned within the group.

[0059] When initially identifying abnormal battery cells, the criteria typically include factors such as the increase in internal resistance, the frequency of internal resistance values ​​exceeding the reference range, and the duration of abnormal fluctuations. These factors help identify battery cells whose performance is gradually deteriorating, or those that suddenly fail under certain special operating conditions.

[0060] The key step in generating a preliminary set of abnormal battery cells is to group these cells together after identifying those with abnormal internal resistance, facilitating subsequent fault analysis, monitoring, and maintenance. This set includes not only cells with abnormal internal resistance but may also contain other potentially faulty cells identified through anomaly detection models.

[0061] During implementation, the internal resistance data of each battery cell is continuously monitored first through a high-frequency data acquisition module. This data is acquired in real time by sensors and uploaded to the battery management system for processing. In the data processing stage, the internal resistance data of each battery cell is first preprocessed, including removing abnormal noise to ensure data accuracy.

[0062] Next, the system generates a dynamic baseline curve based on the internal resistance data of each battery cell within a specific time window. This baseline curve is typically constructed using a sliding window algorithm to smooth the data, ensuring that fluctuations in historical data do not significantly impact the baseline curve. By calculating the mean and standard deviation of the battery's internal resistance within each time window, the generated baseline curve reflects the normal performance fluctuation range of the battery.

[0063] The system then compares the current internal resistance data of each battery cell with the baseline curve, using a set threshold to determine which battery cells have internal resistance exceeding the normal range. These out-of-range cells are marked as abnormal cells and included in a preliminary set of abnormal cells. At this point, a data-driven anomaly detection algorithm may further analyze internal resistance fluctuation patterns to identify battery cells exhibiting progressive degradation or sudden failure characteristics.

[0064] Example Explanation: In a data center battery management system, changes in battery internal resistance are a key factor in assessing battery health. When a battery's internal resistance exceeds a preset benchmark value, the system automatically flags it as an abnormal battery cell and initiates further analysis. This method allows data center operations teams to monitor battery health in real time, promptly identify degraded batteries, and take preventative measures, such as maintenance or replacement, to ensure the UPS system maintains a reliable power supply at all times.

[0065] In medical and healthcare devices, battery stability is crucial for the continuous operation of the equipment. By using dynamic benchmark analysis of internal resistance data, the system can promptly identify battery cells with potential failure risks, thereby preventing the medical device from interrupting operation due to battery failure and ensuring the normal operation of the device at critical moments.

[0066] In the financial sector, critical institutions such as banks and insurance companies require UPS batteries that can operate stably over the long term. Through real-time monitoring and dynamic benchmarking, financial institutions can identify problematic battery cells in advance, enabling them to develop appropriate maintenance plans, reduce the risks of sudden downtime, and ensure the continuity of critical services.

[0067] This embodiment, by monitoring the battery's internal resistance changes in real time and generating a dynamic baseline curve, can more accurately identify battery cells that may fail due to degradation, overuse, or environmental factors. This helps maintenance personnel to promptly detect and address potentially risky batteries before problems become severe, preventing equipment downtime caused by battery failures and effectively improving the efficiency and reliability of battery management.

[0068] S40, For each battery cell in the preliminary abnormal battery cell set, perform multivariate time series analysis using a time series analysis model to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells.

[0069] In this embodiment, when processing each battery cell in the initial set of abnormal battery cells, a time-series analysis model is used to perform multivariate time-series analysis. This involves jointly analyzing historical data of multiple battery performance parameters (such as internal resistance, voltage, and temperature) to help distinguish and confirm abnormal battery cells. Compared with traditional methods that rely on a single indicator for fault detection, the time-series analysis model can more comprehensively and accurately identify the classification of abnormal battery cells.

[0070] First, it is necessary to extract multi-dimensional historical data for each battery cell from the initial set of abnormal battery cells. This historical data includes changes in internal resistance, charging and discharging status, voltage fluctuations, and temperature changes. This data is typically collected through a real-time monitoring system, and the data from each battery cell forms a multi-dimensional time-series data sequence. These historical data sequences are then input into a time-series analysis model. By learning and analyzing the dynamic change patterns of this data, the model can accurately determine whether a battery cell is in a normal, degraded, or failed state.

[0071] The core function of time series analysis models is to classify data based on time series data. Especially when processing historical data of multiple battery cells, they can effectively identify battery cells that exhibit abnormal fluctuations or long-term degradation trends. Time series models, such as Long Short-Term Memory (LSTM) networks or Convolutional Neural Networks (CNNs), can capture long-term dependencies and short-term fluctuation patterns in time series data, thereby distinguishing abnormal states caused by the combined effects of multiple factors such as internal resistance, temperature, and voltage.

[0072] During model training, labeled normal and abnormal battery cell data are used. The trained model can predict the future health status of battery cells based on real-time data. By analyzing the time-series data of each battery cell, the model classifies them into "normal," "degraded," or "failed" categories, ultimately forming a list of confirmed abnormal battery cells.

[0073] During implementation, the historical operating data of each battery cell is first obtained from the initial set of abnormal battery cells, particularly information on internal resistance changes, voltage fluctuations, and temperature changes. This data is acquired in real time through the data acquisition module of the Battery Management System (BMS) and cleaned and standardized using preprocessing algorithms to ensure data quality. Next, this processed time-series data is input into a time-series analysis model. Time-series analysis models, such as LSTM networks, can model historical data to capture long-term and short-term dependencies and abnormal fluctuation patterns in the time series. During model training, a portion of historical data (including battery data labeled "normal" and "abnormal") is typically used for supervised learning, and the loss function is optimized to learn the patterns in the data. The trained time-series analysis model can be used for real-time battery cell status classification. Whenever a battery cell's data is updated, the system automatically uses the time-series analysis model to classify it as "normal" or "abnormal." The analyzed status of each battery cell is added to the list of confirmed abnormal battery cells, thus helping the operations team make further processing decisions.

[0074] Example Description: In data center UPS battery management, data such as battery internal resistance, temperature, and voltage are collected in real time by a monitoring system and transmitted to the battery management system. Historical data for each battery cell is input into a time-series analysis model for training. The model can classify the battery's health status based on this data. For example, when both internal resistance and temperature values ​​fluctuate abnormally, the model can identify that these fluctuations are caused by battery degradation, mark the battery as abnormal, and promptly add it to the list of confirmed abnormal battery cells. In this way, the data center operations team can identify degraded batteries in advance and perform maintenance or replacement before serious battery failures occur, thereby avoiding sudden downtime and ensuring the efficient operation of the data center.

[0075] In the field of medical devices, batteries are a critical power source, and ensuring their proper functioning is paramount. Through multivariate time-series analysis, medical device administrators can identify potential battery failures in advance and replace problematic batteries promptly, preventing equipment downtime due to battery issues and ensuring a continuous power supply for patients during treatment.

[0076] In the financial sector, battery management systems in financial institutions such as banks and insurance companies face similar challenges. Accurate battery condition classification allows for early detection of battery degradation, preventing sudden power outages and ensuring the continued operation of critical business activities.

[0077] This embodiment utilizes a multivariate time-series analysis model to accurately classify abnormal battery cells, significantly improving the identification accuracy. Compared to traditional detection methods based on a single internal resistance index or manual thresholds, it can not only identify battery cells that are deteriorating early but have not yet reached the warning threshold, but also promptly and accurately identify problematic batteries when their performance deviates in multiple dimensions. This allows the maintenance team to address problematic batteries before they truly fail, effectively extending battery life, reducing downtime risks, and improving the intelligence level of battery management.

[0078] S50, Based on the confirmed abnormal battery cell list and the preset maintenance knowledge base, generate a battery cell maintenance plan table.

[0079] In this embodiment, generating a battery cell maintenance plan based on the confirmed abnormal battery cell list and a preset maintenance knowledge base is a key operation to ensure timely maintenance and replacement of battery cells. The core of this step lies in combining the battery cell health status data with the existing maintenance strategy library to develop a structured and executable maintenance plan.

[0080] First, confirm that the health status of each battery cell in the abnormal battery cell list has been categorized and identified as "abnormal" or "failed". For each battery cell marked as abnormal, it is necessary to first associate its static attribute data, including information such as brand, model, capacity, and service life. This static attribute data provides detailed background information for the battery cell, which helps determine the maintenance needs and priorities of each battery cell.

[0081] Next, the system performs a maintenance strategy matching operation by combining existing maintenance strategies from the preset maintenance knowledge base. The preset maintenance knowledge base typically includes standard operating procedures (SOPs) for battery maintenance, maintenance cycles, processing priorities, and handling measures for various faults. For example, when the internal resistance of some battery cells changes beyond a set threshold, the system can find suitable handling measures from the maintenance knowledge base, such as deep discharge, adjusting the charging strategy, or replacing the battery.

[0082] Based on the matched maintenance strategy, the system then fills in the corresponding data fields according to the preset maintenance plan template to generate a battery cell maintenance plan. The maintenance plan typically includes information such as the identifier of each battery cell, its health status, the maintenance measures to be performed, maintenance priority, execution time, and responsible person. This data not only helps optimize the battery cell maintenance process but also improves work efficiency, ensuring timely handling of problematic battery cells.

[0083] In the implementation process, the system first retrieves the identifier of each battery cell from the list of confirmed abnormal battery cells and then retrieves its static attribute data from the database based on these identifiers. This static attribute data includes, but is not limited to, basic information such as the battery's brand, model, capacity, and service life. This information is crucial for subsequently determining the battery's maintenance priority and applicable strategies. Next, the system automatically matches a maintenance strategy for each abnormal battery cell using a pre-set maintenance knowledge base. Maintenance strategies include various possible handling measures, such as adjusting the charging mode, deep discharge, heating, discharge cycling, and battery replacement. These strategies in the maintenance knowledge base are closely related to the battery cell's static attributes and abnormal state. For example, if the battery has a high internal resistance and has undergone many charge cycles, the system may recommend replacing the battery. Once a suitable maintenance strategy is matched, the system fills in the data according to a pre-set maintenance plan template to generate a battery cell maintenance plan. This plan includes: the battery cell's identifier, abnormality type, maintenance priority, maintenance measures to be implemented, estimated execution time, and responsible person. Each data item in the plan is automatically filled in based on the actual situation of the battery cell and sorted according to the priority and execution time of the maintenance strategy. In this way, battery managers can clearly see which battery cells need maintenance, ensuring that each battery cell receives timely and appropriate treatment.

[0084] Example Description: In data center UPS battery management, data on changes in battery cell internal resistance, temperature fluctuations, and voltage variations are monitored and stored in real time by the Battery Management System (BMS). When a battery cell's internal resistance exceeds a threshold, the system automatically identifies the battery as "abnormal" and retrieves corresponding maintenance strategies from a pre-defined maintenance knowledge base based on static attribute data such as the battery's brand, model, capacity, and service life. For some older batteries, the system may recommend deep discharge, while for batteries with severely elevated internal resistance, the system will recommend replacement. Finally, the system automatically generates a maintenance plan, listing all battery cells requiring maintenance, along with corresponding maintenance measures and execution times, ensuring that maintenance personnel can handle each abnormal battery cell promptly and accurately.

[0085] In battery management for medical devices, battery internal resistance changes and temperature data are collected in real time via sensors. If an abnormal internal resistance is detected, the system generates a maintenance plan based on static attributes such as the battery's age and operating environment. For example, for spare batteries that have not been used for a long time, the system may recommend periodic charge-discharge cycles to restore their capacity. The generated maintenance plan helps equipment managers perform timely battery maintenance, preventing equipment downtime due to battery failure and ensuring the normal operation of the equipment.

[0086] In the financial sector, particularly in backup battery management at banks and other locations, battery health directly impacts the stability of critical systems. Through automated maintenance plan generation, the system can effectively monitor batteries and provide timely maintenance recommendations, preventing service interruptions due to battery failure and ensuring business continuity.

[0087] This embodiment automatically generates a battery cell maintenance plan by combining abnormal battery cell states with strategies from a maintenance knowledge base, thereby reducing manual intervention and ensuring timely maintenance and replacement of battery cells. The ability to automatically generate maintenance plans based on battery health improves the accuracy and efficiency of battery management while reducing unnecessary maintenance work. The structured output of the maintenance plan also makes battery management more standardized and efficient, significantly reducing the impact of battery failure on equipment and systems.

[0088] This invention relates to the field of data processing technology and can be applied to business scenarios such as battery management, fintech, and healthcare. It discloses a method, apparatus, device, and medium for detecting abnormal battery cells, comprising: acquiring static attribute data and operating status parameters of multiple battery cells; generating a homogeneous battery cell group through dynamic clustering; performing dynamic benchmark analysis based on internal resistance data within the homogeneous battery cell group to initially identify abnormal battery cells and generate a preliminary set of abnormal battery cells; performing multivariate time-series analysis on each battery cell in the preliminary set of abnormal battery cells using a time-series analysis model to generate a list of confirmed abnormal battery cells; and generating a battery cell maintenance plan based on the list of confirmed abnormal battery cells and a preset maintenance knowledge base. This invention, by combining dynamic benchmark analysis and a time-series analysis model, accurately identifies abnormal states of battery cells, avoiding misjudgments caused by fixed thresholds in traditional methods. It can promptly detect potentially problematic batteries, improving the efficiency and accuracy of battery monitoring systems and ensuring battery reliability at critical moments.

[0089] In one embodiment, step S10 includes:

[0090] S101 collects operating status parameters for each battery cell, including internal resistance, voltage, and temperature.

[0091] S102, Obtain static attribute data for each battery cell, including brand, model, capacity, and service life;

[0092] S103, Perform a data cleaning operation on the running status parameters to generate cleaned running status parameters;

[0093] S104, Perform a time alignment operation on the cleaned operating status parameters to generate time-aligned operating data;

[0094] S105, Perform a statistical aggregation operation on the time-aligned running data to generate running status aggregated data;

[0095] S106, Store the aggregated running status data and the static attribute data.

[0096] In this embodiment, acquiring and processing the battery's static attribute data and operational status parameters is a fundamental step in ensuring real-time monitoring of the battery's health status. Specifically, the static attribute data and operational status parameters of the battery cell are typically collected in real time by multiple sensors and data acquisition devices. Static attribute data includes battery information that does not change over time, such as brand, model, capacity, and service life, while operational status parameters, such as internal resistance, voltage, and temperature, are dynamic manifestations of the battery's health status.

[0097] First, for each battery cell, the acquired operating parameters include internal resistance, voltage, and temperature. Internal resistance is a crucial indicator of battery health, and its changes directly impact battery performance and lifespan. Voltage and temperature affect the battery's charge / discharge efficiency and safety. Acquiring these parameters is typically accomplished by sensors built into the battery cell and an external monitoring system, with data transmitted and stored via IoT devices.

[0098] Next, the acquired raw operating parameters need to undergo data cleaning. This operation primarily aims to remove noise, missing data, or outliers that may exist during the data acquisition process, ensuring the accuracy and consistency of the data. The cleaned data forms the basis for further analysis and processing, helping to improve the reliability of the analysis results.

[0099] After cleaning, the data typically requires time alignment. The purpose of time alignment is to ensure that data collected from different battery cells or sensors are time-aligned, enabling effective comparison and aggregation of parameter data from different devices or points in time. Time-aligned data provides a unified temporal basis for subsequent statistical analysis and data mining.

[0100] Statistical aggregation is a crucial step in summarizing and analyzing time-aligned data. During this process, statistical processing of parameters such as internal resistance, voltage, and temperature generates aggregated data for each battery cell. This aggregated data includes hourly maximum, average, and minimum values, reflecting the battery cell's health and performance over different time periods. The maximum, average, and minimum values ​​respectively provide the extreme range, average level, and lowest point of battery performance fluctuations, offering strong support for battery maintenance and replacement decisions.

[0101] Finally, all generated runtime status aggregate data and static attribute data will be stored. Data storage can utilize traditional database systems or cloud-based storage systems to facilitate subsequent queries, analysis, and historical data comparison. This data storage will serve as a crucial basis for battery monitoring and maintenance decisions.

[0102] This embodiment achieves accurate monitoring of battery cell health status by acquiring, cleaning, aligning, and aggregating static attribute data and operational status parameters of multiple battery cells. This series of operations effectively eliminates noise and inconsistencies in the data, ensuring the reliability and accuracy of the analysis results. Through time alignment and statistical aggregation, not only can the current state of the battery be comprehensively understood, but also subtle fluctuations in battery performance can be detected, helping to identify potential failure risks in a timely manner. In addition, storing the cleaned and aggregated data provides a comprehensive historical data record, facilitating long-term monitoring and optimization of the battery management system, improving battery utilization efficiency and extending battery life, ultimately reducing sudden power outages caused by battery failure and ensuring system reliability.

[0103] In one embodiment, step S20 above includes:

[0104] S201, Perform category attribute encoding and numerical attribute standardization processing on the static attribute data to generate standardized static attribute data;

[0105] S202, Perform a feature selection operation on the standardized static attribute data to determine clustering feature data;

[0106] S203, Perform distance optimization grouping operation on the clustering feature data to generate initial battery cell groups;

[0107] S204, Perform an attribute consistency verification operation on the initial battery cell group to generate a homogeneous battery cell group;

[0108] S205, record the division results of the homogeneous battery cell group.

[0109] In this embodiment, multiple battery cells are dynamically clustered based on static attribute data to group battery cells with similar performance, thereby improving the efficiency and accuracy of battery management. The core of the clustering operation is to process the static attributes of the battery cells so that battery cells with similar characteristics can be grouped together, facilitating subsequent monitoring, analysis, and maintenance.

[0110] First, categorical attribute encoding and numerical attribute standardization are performed on the static attribute data. Categorical attribute encoding refers to converting the categorical characteristics of battery cells (such as brand, model, etc.) into numerical data that can be processed by a computer. For example, the brand attribute can be implemented by assigning unique code values ​​to different brands, while attributes such as model and capacity can be directly quantified. Numerical attribute standardization aims to eliminate the differences between attributes caused by different units of measurement, ensuring that all attributes are compared under the same unit, thereby avoiding interference with the clustering results. For example, capacity is usually measured in mAh (milliampere-hours), while voltage is measured in volts (V). Standardization will unify the value range of these numerical attributes.

[0111] Secondly, feature selection is performed on the standardized static attribute data. Feature selection aims to filter out the features most relevant to clustering from all attribute data, thus improving clustering accuracy and reducing computational complexity. Feature selection analyzes the correlation between each attribute and the target variable, or uses algorithms such as Principal Component Analysis (PCA) to remove redundant attributes, ensuring that clustering relies only on the most meaningful data.

[0112] Next, a distance-optimized grouping operation is performed on the selected clustering feature data. The purpose of this operation is to group the battery cells based on certain distance metrics (such as Euclidean distance or Manhattan distance). The key to the distance-optimized grouping operation is to determine which battery cells should be grouped together by calculating the similarity between different battery cells. This process classifies battery cells into appropriate groups based on their characteristics, thus ensuring that the clustering results of the battery cells are practically meaningful.

[0113] Next, a property consistency verification operation is performed on the generated initial battery cell groups. This step further verifies the initial grouping results, ensuring that the battery cells within each group are consistent in properties. For example, within the same battery cell group, properties such as capacity and voltage should be similar. If battery cells within a group show significant differences in these properties, the grouping needs to be readjusted until the properties of the battery cells within the group meet the consistency requirements.

[0114] Finally, the classification results of the homogeneous battery cell groups are recorded. Recording these classification results provides a traceable data foundation for subsequent analysis, monitoring, and maintenance. By recording the group to which each battery cell belongs, the battery pack can be tracked in real time during subsequent monitoring, and battery cells exhibiting abnormal behavior can be identified promptly.

[0115] In different implementations, the above clustering operations can be achieved using different techniques. For example, for static attribute data of battery cells, the K-means clustering algorithm in machine learning can be used for distance-optimized grouping. The K-means algorithm divides data points into K clusters through iterative optimization, with data points within each cluster being as similar as possible. This algorithm shows good performance in dynamic clustering of battery cells, and is particularly suitable for processing large amounts of battery cell data. Furthermore, feature selection operations can be implemented using various methods. In some scenarios, feature selection methods based on correlation analysis (such as Pearson correlation coefficient or information gain) can effectively identify and filter features closely related to the battery health status, while ignoring irrelevant or redundant features. In other scenarios, dimensionality reduction methods such as PCA are used to reduce the dimensionality of the feature space, thereby improving the efficiency of subsequent clustering operations.

[0116] In practical applications, the choice of clustering algorithm and the implementation of feature selection can be adjusted according to the different characteristics of battery cells. For example, some high-end battery monitoring systems may use more complex clustering algorithms (such as hierarchical clustering or DBSCAN) to accommodate more diverse battery cell characteristics. Feature selection operations can also be fine-tuned according to the specific equipment performance to ensure a more accurate clustering process.

[0117] This embodiment effectively groups battery cells according to similar characteristics through dynamic clustering based on static attribute data, providing a solid foundation for subsequent monitoring, maintenance, and optimization. Battery cells within each group possess similar performance characteristics, allowing maintenance personnel to take more precise management measures based on group characteristics. This clustering operation significantly improves battery management efficiency, reduces misjudgments and delayed responses caused by the increased complexity and cumbersome data processing in traditional methods, thereby enhancing the response speed and accuracy of battery monitoring and maintenance.

[0118] In one embodiment, step S30 above includes:

[0119] S301, Based on the internal resistance data in the operating status parameters, perform a benchmark determination operation on the homogeneous battery cell group to generate dynamic benchmark curve data;

[0120] S302, performs a standard score determination operation on the internal resistance data of each battery cell to generate an internal resistance standard score sequence;

[0121] S303, Perform a continuous deviation detection operation on the internal resistance standard fraction sequence to generate a marker for the continuously deviating battery cell;

[0122] S304, Perform an anomaly identification operation on the continuously deviating mark from the battery cell to generate a preliminary set of abnormal battery cells;

[0123] S305, based on the internal resistance data in the newly acquired operating status parameters, update the dynamic reference curve data.

[0124] In this embodiment, dynamic benchmark analysis of homogeneous battery cell packs based on internal resistance data enables early anomaly identification, ensuring effective monitoring of the battery pack's health status. Changes in internal resistance are a crucial indicator of battery degradation; when a battery cell's internal resistance exceeds the normal range, it often signifies performance degradation and even a risk of failure. Therefore, dynamic benchmark analysis based on internal resistance data allows for timely identification of potential battery malfunctions.

[0125] First, a benchmark determination operation is performed on the homogeneous battery cell group based on the internal resistance data in the operating status parameters. The purpose of this operation is to generate dynamic benchmark curve data within the battery cell group. These benchmark curves reflect the changing trend of the internal resistance of the battery cells under normal operating conditions. During the generation of dynamic benchmark curves, the battery's internal resistance data is integrated and processed based on its historical performance and current state, thus providing a reasonable benchmark for subsequent anomaly identification. In the benchmark determination process, methods such as regression analysis or smoothing techniques can be used to ensure that the benchmark curves accurately reflect the normal operating range of the battery cells.

[0126] Next, a standard score determination operation is performed on the internal resistance data of each battery cell. This operation aims to convert the internal resistance data of each battery cell into a standard score (Z-score), facilitating the comparison of the internal resistance performance of different battery cells. The standard score is calculated by comparing the internal resistance data with the average internal resistance within the battery pack and measuring the degree of deviation based on the standard deviation. This step helps eliminate differences caused by data fluctuations between battery cells, making the degree of fluctuation in internal resistance data more significant. The generation of the standard score sequence plays a crucial role in subsequent anomaly identification.

[0127] After generating the standard score sequence, a persistent deviation detection operation is performed on these sequences. The core purpose of persistent deviation detection is to identify battery cells whose internal resistance standard score consistently exceeds a preset threshold over a period of time. This operation can identify battery cells that may be experiencing early degradation, as significant deviations in internal resistance are often signals of battery degradation or failure. Persistent deviation detection is based on the time series of internal resistance data, ensuring that the deviation is persistent and stable, rather than a short-term fluctuation.

[0128] Subsequently, an anomaly identification process is performed on battery cells marked as continuously deviating. The purpose of this step is to further confirm whether the battery cell is in a faulty or abnormal state. The anomaly identification process can combine historical data and battery cell performance metrics, using algorithms to evaluate the deviation and determine whether it constitutes a genuine anomaly. Common methods include judgments based on statistical models, machine learning models, or rule engines. This process generates a preliminary set of abnormal battery cells, providing a clear basis for battery management and subsequent maintenance.

[0129] Finally, the dynamic baseline curve data is updated based on the internal resistance data from the newly acquired operating status parameters. As the battery cells are used, the internal resistance value may change, therefore the baseline curve needs to be updated periodically to ensure the real-time nature and accuracy of the baseline data. The updated baseline curve will compare with the new internal resistance data, helping to promptly identify new abnormal trends. The update process can be performed periodically or when the state of the battery cells changes significantly.

[0130] This embodiment utilizes dynamic benchmark analysis and anomaly identification based on internal resistance data. This technical solution effectively improves the accuracy and efficiency of battery cell management. By monitoring changes in internal resistance in real time and comparing them with a dynamic benchmark, potentially faulty or degraded battery cells can be identified early, preventing them from affecting the overall performance of the battery pack. This significantly reduces the lag in traditional manual monitoring and accurately identifies battery cells requiring maintenance, reducing unnecessary maintenance costs. Furthermore, by regularly updating the benchmark curve, continuous tracking of battery status is ensured, further improving the system's reliability and stability.

[0131] In one embodiment, step S304 includes:

[0132] S3041, for each mark continuously deviating from the battery cell, obtain its internal resistance standard score sequence;

[0133] S3042, Perform a continuous deviation segment identification operation on the internal resistance standard fraction sequence to generate a continuous deviation segment set;

[0134] S3043, Perform a duration determination operation for each consecutive deviation segment to generate a segment duration value;

[0135] S3044, Perform an abnormality persistence determination operation based on the duration value of the segment, generate a persistent abnormal state determination result, and record the generation timestamp of the persistent abnormal state determination result;

[0136] S3045, Perform an anomaly confirmation operation on the continuous abnormal state judgment result to generate a confirmed abnormal battery unit;

[0137] S3046, Summarize all confirmed abnormal battery cells, generate a preliminary abnormal battery cell set, and record the generation timestamp of the preliminary abnormal battery cell set.

[0138] In this embodiment, by using continuously deviating markers, potentially abnormal battery cells can be effectively identified, preventing battery degradation from affecting the overall system. Changes in internal resistance often reflect the health of the battery, especially during long-term use. Therefore, by performing anomaly identification operations on battery cells with continuously deviating markers, degraded batteries can be detected in a timely manner, ensuring the health of the battery pack.

[0139] First, for each battery cell marked as exhibiting a persistent deviation, its internal resistance standard score sequence is obtained. The purpose of this operation is to obtain a standardized sequence of the battery cell's internal resistance data, facilitating comparison with the normal range. The internal resistance standard score sequence represents the degree of change in the battery's internal resistance; a higher standard score indicates a greater deviation from the normal value, thus allowing for more precise identification of which battery cells exhibit abnormal internal resistance trends.

[0140] Next, a continuous deviation segment identification operation is performed on the internal resistance standard score sequence. This operation aims to identify segments in the internal resistance standard score sequence that continuously exceed the normal range (i.e., the threshold). By identifying these segments, the duration of the battery's abnormality can be further analyzed. For example, when the standard score of internal resistance continuously exceeds the threshold, it indicates that the battery's health may be deteriorating and requires special attention. This step helps filter out occasional, short-lived fluctuations, thus ensuring that only anomalies with long-term trends are considered.

[0141] After generating the set of consecutive deviation segments, a duration determination operation needs to be performed on each segment. This operation calculates the duration value of each segment, typically determined by the difference between the segment's start and end times. By quantifying the duration of each deviation segment, the time span of each abnormal trend can be more clearly understood, and it can be further determined whether it meets the criteria for a long-term anomaly. The duration value is a crucial basis for judging whether an anomaly is persistent; a prolonged period of continuous deviation often indicates a significant risk of battery failure.

[0142] Based on the duration value of the segment, an abnormality persistence determination operation is performed. The purpose of this operation is to determine whether these deviation segments meet the criteria for long-term abnormality. Typically, the duration value is compared with a preset threshold. If the duration exceeds the threshold, it is judged as an abnormality persistence; otherwise, it is considered a transient fluctuation. In this way, battery cells with genuine abnormalities can be identified more accurately, avoiding the misjudgment of temporary fluctuations as malfunctions.

[0143] After determining the persistence of anomalies, an anomaly confirmation operation needs to be performed on the results. The purpose of this confirmation is to further verify which battery cells are indeed abnormal, rather than being due to errors caused by external factors. For example, some battery cells may exhibit significant fluctuations under specific external conditions, but after anomaly confirmation, it may be found that these fluctuations are sporadic, rather than indicating battery degradation. This operation is verified by combining historical data, diagnostic algorithms, or other rule engines to ensure a high degree of accuracy in identifying the final confirmed abnormal battery cells.

[0144] Finally, all confirmed abnormal battery cells will be aggregated to generate a preliminary set of abnormal battery cells, and the generation timestamp will be recorded. By bringing together all confirmed abnormal battery cells, a complete list of abnormal cells can be formed, facilitating subsequent processing and analysis. Furthermore, recording the timestamps is crucial for subsequent trend analysis and historical data retrospective analysis, helping maintenance personnel trace the occurrence time of abnormal battery cells and analyze their changing trends.

[0145] This embodiment accurately identifies abnormal conditions within battery cells, particularly those that have been in an abnormal state for an extended period, by performing anomaly identification operations on cells whose markers continuously deviate from their designated positions. This avoids the shortcomings of traditional technologies that rely solely on threshold alarms, filtering out short-term fluctuations and ensuring that only genuine anomalies are addressed. Furthermore, the recording of timestamps and determination of duration improve the reliability and accuracy of anomaly detection, providing strong support for battery health monitoring and subsequent maintenance. This method not only enhances the accuracy of battery cell anomaly detection but also effectively reduces false alarms, ensuring the precise implementation of maintenance plans and improving data center operational efficiency and battery system stability.

[0146] In one embodiment, step S40 above includes:

[0147] S401, For each battery cell in the preliminary abnormal battery cell set, extract the internal resistance change characteristics and voltage fluctuation characteristics from the operating state parameters;

[0148] S402, Based on the internal resistance data and temperature data in the operating status parameters, perform a temperature correlation analysis operation to generate temperature correlation results;

[0149] S403, using a time series analysis model, an anomaly classification operation is performed based on the internal resistance change characteristics, the voltage fluctuation characteristics, and the temperature correlation results to generate anomaly classification results;

[0150] S404, Perform a false alarm filtering operation on the abnormal classification results, remove transient interference abnormal classification results, and generate a list of confirmed abnormal battery cells.

[0151] In this embodiment, accurate identification of abnormal battery cells is crucial for ensuring the stability of the battery pack. By combining multiple features from the operating state parameters, especially internal resistance changes, voltage fluctuations, and temperature correlations, anomaly classification can be performed more accurately. The core of this process lies in processing these features through a time-series analysis model to identify truly abnormal battery cells and avoid misjudgments caused by short-term fluctuations or external factors.

[0152] First, for each battery cell in the initial set of abnormal battery cells, the internal resistance variation characteristics and voltage fluctuation characteristics are extracted from the operating state parameters. Internal resistance variation characteristics accurately reflect the battery's health status; an increase in internal resistance usually indicates a decline or degradation in battery performance. Voltage fluctuation characteristics provide information on battery performance changes under different loads. By extracting these characteristics, necessary data support can be provided for subsequent classification, ensuring that the battery performance characteristics involved in the classification process are fully considered.

[0153] Next, based on the internal resistance and temperature data in the operating status parameters, a temperature correlation analysis is performed to generate temperature correlation results. Temperature changes typically have a significant impact on battery performance, especially the relationship between internal resistance and temperature, which can reflect the battery's operating state under different temperature environments. By performing correlation analysis, the relationship between internal resistance and temperature can be quantified, providing more dimensions of support for anomaly classification and thus improving the accuracy of anomaly detection.

[0154] The use of time-series analysis models is central to the anomaly classification process. In this step, based on the extracted internal resistance variation characteristics, voltage fluctuation characteristics, and temperature correlation results, the time-series analysis model performs an in-depth analysis of the battery cell behavior. Time-series analysis models (such as LSTM and GRU deep learning models) can handle the correlation and dependency of time-series data. Through the analysis of historical data, the model can identify potential anomaly patterns and classify battery cells. The model training process ensures the accuracy and reliability of battery cell classification, enabling the differentiation between truly abnormal battery cells and normal battery cells.

[0155] Perform false alarm filtering on anomaly classification results. The purpose of false alarm filtering is to remove misclassification results caused by transient interference or short-term fluctuations. Fluctuations in the battery system may be caused by changes in the external environment (such as temperature fluctuations) or other factors not related to the battery itself; these errors need to be filtered out during anomaly identification. Through false alarm filtering, the system can more accurately identify long-term stable anomalies, reduce the need for human intervention, and improve the level of automation in detection.

[0156] False alarm filtering can be implemented through various means, aiming to effectively distinguish between false alarms caused by short-term fluctuations or external disturbances and genuine abnormal states. First, the system needs to set a threshold or standard to identify the characteristics of short-term fluctuations. Typically, these short-term fluctuations are short in duration, small in amplitude, and usually recover quickly to normal levels within a period of time. These fluctuations may be caused by external environmental factors (such as sudden changes in temperature, humidity, or instantaneous fluctuations during battery charging). To achieve accurate false alarm filtering, the system compares the anomaly classification results with historical battery data, especially with past fluctuation patterns. By analyzing the historical data of the battery cells, the system can determine which anomalies are caused by external environmental or non-battery factors. The system can use a sliding window technique to analyze fluctuations over a period of time. If the fluctuation duration is shorter than a preset threshold and the fluctuation recovery time is rapid, it can be judged as a false alarm, and these false alarm data can be filtered out. In addition, combining multi-factor analysis also helps reduce the generation of false alarms. For example, if multiple operating parameters such as internal resistance, voltage, and temperature fluctuate simultaneously, and these fluctuations are not related to the battery's health condition (such as aging or wear), the system can classify them as transient disturbances rather than genuine anomalies. These data will be marked as "non-abnormal," ensuring that only long-term stable abnormal states are identified, thus avoiding false alarms. In short, the key to false alarm filtering lies in comparing and analyzing historical data, fluctuation duration, and external environmental factors to remove misclassifications caused by short-term fluctuations or transient disturbances. This ensures the system can accurately identify long-term stable and truly abnormal battery cells, thereby improving the accuracy and reliability of the entire detection process.

[0157] Finally, a list of confirmed faulty battery cells is generated. This list contains all battery cells identified as faulty, requiring subsequent maintenance, inspection, or replacement. By generating this list, the system provides battery managers with clear operational guidelines, helping them focus resources on the most critical battery cells, thereby optimizing maintenance and extending the battery pack's lifespan.

[0158] This embodiment utilizes a time-series analysis model to comprehensively analyze the correlation between internal resistance changes, voltage fluctuations, and temperature, enabling more accurate identification of abnormal battery cells, especially in the early degradation stage. When internal resistance data, temperature fluctuations, and voltage changes simultaneously indicate changes in battery health, the system can efficiently filter out battery cells requiring attention. Furthermore, a false alarm filtering mechanism reduces misjudgments caused by short-term fluctuations, improving the accuracy of anomaly identification, thereby effectively extending battery life and reducing maintenance costs.

[0159] In one embodiment, step S50 includes:

[0160] S501, For each battery cell in the confirmed abnormal battery cell list, associate its static attribute data to generate a detailed list of abnormal batteries;

[0161] S502, based on the preset maintenance knowledge base, perform a maintenance strategy matching operation on the detailed list of abnormal batteries to generate a maintenance strategy matching result;

[0162] S503, according to the preset maintenance plan template, perform a plan filling operation on the maintenance strategy matching result to generate a battery cell maintenance plan.

[0163] In this embodiment, generating a maintenance plan is a crucial step in ensuring timely and effective maintenance of the battery cells. This process relies on identifying a list of abnormal battery cells and an existing maintenance knowledge base to provide personalized maintenance strategies and ultimately create a clear maintenance plan.

[0164] First, each battery cell in the confirmed abnormal battery cell list is associated with its static attribute data. Static attribute data refers to basic information recorded at the initial stage of battery manufacturing or deployment. This typically includes brand information (to distinguish product models from different manufacturers), capacity information (to differentiate product specifications), rated capacity parameters, and service life information (to describe the cumulative working cycles or time since the battery was put into use). All of this information originates from data records generated by the system during battery deployment or routine inspections. The association operation involves retrieving and binding the abnormal objects in the confirmed abnormal battery cell list with the static attribute data in the database based on the unique identifier of each battery cell, forming a detailed list of abnormal batteries. This list not only includes the basic attribute information of each abnormal battery but may also include extended information such as the abnormality type, abnormality level, detection time, and historical maintenance records, facilitating subsequent maintenance decision-making.

[0165] After obtaining a detailed list of abnormal batteries, a maintenance strategy matching operation is performed based on a pre-defined maintenance knowledge base. The maintenance knowledge base refers to a pre-established set of standardized rules in the system, covering standard maintenance measures and strategies for different brands, models, capacities, battery types, and abnormal conditions. This knowledge base typically draws from multi-dimensional information sources such as technical manuals provided by equipment manufacturers, industry standards, historical maintenance data, and expert experience rules. The maintenance strategy matching operation compares the information in the detailed list of abnormal batteries with the maintenance rules in the maintenance knowledge base through logical judgment and rule matching to determine specific maintenance measures. For example, if a battery model exhibits abnormal internal resistance during testing and its service life is nearing its limit, the system can automatically match a strategy recommending replacement; if the battery has slight voltage fluctuations and other parameters are within a controllable range, the system will match a suggestion for periodic observation or adjustment of the charge / discharge strategy. The matching results form a maintenance strategy matching result data structure, containing specific maintenance measures, priority arrangements, and operational suggestions for each abnormal battery cell.

[0166] After strategy matching is completed, the system performs a plan population operation based on a preset maintenance plan template to generate a battery unit maintenance plan. The maintenance plan template is a structured data table or electronic document framework that defines various necessary information fields for maintenance tasks, such as battery unit number, anomaly type, maintenance strategy, operation steps, execution time nodes, responsible personnel information, and execution status. During the plan population operation, the system, based on the maintenance strategy matching results, fills in the corresponding fields of the maintenance plan for each abnormal battery unit, ensuring complete information, clear logic, and a consistent structure. The final battery unit maintenance plan has clear task arrangements, operation specifications, and execution standards, facilitating efficient execution of maintenance tasks by the operations and maintenance team, reducing the risk of omissions, and improving overall management efficiency.

[0167] Example Description: In the field of battery management, addressing the operation and maintenance needs of backup lead-acid batteries for data center UPS systems, a data center server room is configured with a backup battery bank containing 12,000 lead-acid batteries. The system continuously collects the operating status parameters of all battery cells, specifically including the internal resistance, voltage, and temperature data of each cell. Simultaneously, it acquires the static attribute data of each battery cell, including information such as brand, model, rated capacity, and service life. The operating status parameters are transmitted in real-time through the data center's environmental monitoring system, while the static attribute data is automatically read from the equipment files in the maintenance system.

[0168] After data acquisition, the system performs data cleaning on the operating status parameters, removing outliers and failure points from the data transmission process to generate continuous and valid cleaned operating status parameters. For the cleaned data, the system performs time alignment processing based on the battery operation monitoring timeline, ensuring that the monitoring data of all battery cells are analyzed synchronously with the same time base. The time-aligned data is then further calculated hourly for each battery cell, including the maximum, average, and minimum values ​​of internal resistance, voltage, and temperature, forming aggregated operating status data. This aggregated data, along with the static attribute data, is stored in the battery status management database.

[0169] For the stored data, the system performs categorical attribute encoding on the static attribute data, converting brand and model into numerical category indexes. Simultaneously, it standardizes the capacity and service life data to ensure that all feature parameters are processed on the same numerical scale. After feature selection, the system retains brand, model, capacity, and service life—highly correlated with battery health status—as clustering features. Based on these clustering features, the system uses a distance-optimized grouping method to dynamically cluster all battery cells, automatically grouping cells with high data similarity and consistent battery attributes into initial battery cell groups. Attribute consistency verification is performed on these initial battery cell groups to ensure high consistency in key attributes such as brand, model, and capacity. Upon successful verification, homogeneous battery cell groups are generated, and the division results are recorded in the battery grouping management table.

[0170] Within each homogeneous battery cell group, the system selects the internal resistance data of the corresponding battery, performs a benchmark determination operation, and periodically calculates the mean and standard deviation of internal resistance based on historical data to generate dynamic benchmark curve data, which serves as a comparison benchmark for changes in internal resistance within the group. For the internal resistance data of each battery cell, the system further calculates a series of standard internal resistance scores to measure the degree of deviation of the battery's internal resistance from the benchmark curve at each time point. Through continuous deviation detection, the system automatically identifies battery cells whose standard internal resistance scores continuously exceed a set threshold over multiple consecutive time periods and marks them as continuously deviating battery cells.

[0171] For marked battery cells exhibiting continuous deviation, the system acquires their complete internal resistance standard score sequence, performs continuous deviation segment identification, locates all time periods of continuous deviation, and calculates the duration of each continuous deviation segment. The system compares the duration of each segment with preset abnormality persistence criteria, generates a continuous abnormality state judgment result, and records the timestamp of this judgment result in real time. The system inputs the continuous abnormality state judgment result into the abnormality confirmation operation, cross-validating historical data with the current state to confirm whether it belongs to a long-term internal resistance abnormality. All confirmed abnormal battery cells are automatically aggregated by the system to form a preliminary abnormal battery cell set, and the time point of set generation is recorded.

[0172] For each battery cell in the initial set of abnormal battery cells, the system extracts its operating state parameters and calculates the rate of change of internal resistance and the amplitude of voltage fluctuation as time-series features of the battery. Combining the internal resistance data and temperature data of the battery cell, the system performs temperature correlation analysis, generating temperature correlation results that reflect the coupled change trend of battery internal resistance and temperature. The aforementioned internal resistance change features, voltage fluctuation features, and temperature correlation results are input into the trained time-series analysis model to perform anomaly classification. The system categorizes abnormal batteries to determine whether they are genuinely degraded batteries or false alarm interference batteries. Based on the anomaly classification results, the system further applies a false alarm filtering mechanism, dynamically analyzing parameters such as the duration of the anomaly, fluctuation patterns, and correlation with environmental changes, automatically eliminating false alarms caused by short-term external interference or temperature transients, and finally generating a list of confirmed abnormal battery cells.

[0173] Based on the confirmed list of abnormal battery cells, the system associates the static attribute data of each battery cell to generate a detailed list of abnormal batteries, recording information such as brand, model, capacity, service life, anomaly type, and anomaly occurrence time for each abnormal battery. The system then calls upon the maintenance knowledge base, matching the parameters in the detailed list of abnormal batteries with the maintenance rules within the knowledge base to generate a maintenance strategy matching result. This determines the processing priority and maintenance recommendations for different battery cells, such as deep discharge, parameter adjustment, or battery replacement. Finally, following the maintenance plan template, the system populates the confirmed list of abnormal battery cells and the matching maintenance strategies into a standardized maintenance plan table, generating a battery cell maintenance plan table containing information such as battery number, anomaly type, maintenance measures, operation steps, execution time nodes, and responsible personnel. This plan is provided to the operations and maintenance team as a direct reference for subsequent on-site maintenance. The entire process achieves closed-loop automated management of data center lead-acid batteries, from data acquisition, cluster analysis, anomaly detection, classification confirmation, maintenance decision-making to plan table generation.

[0174] In the healthcare sector, the management of backup lead-acid batteries widely used in hospital backup power systems is crucial. Large hospital central power supply systems typically include multiple UPS backup battery banks. These battery units directly impact the continuous power supply capability of operating rooms, emergency equipment, and intensive care equipment during main power failures. The hospital power management system continuously collects operational status parameters from all backup battery units, including internal resistance, voltage, and temperature data. It also acquires static attribute information for each battery unit, including brand, model, capacity, and service life. Operational status parameters are transmitted in real-time through the hospital equipment monitoring platform, while static attribute information is synchronously imported through the power asset management system.

[0175] During data acquisition, the hospital's power management system first removes outliers from operating status parameters, eliminating invalid data caused by monitoring anomalies or equipment transmission interruptions, and generating a continuous and valid data sequence. The cleaned data is then uniformly aligned to a time base to ensure that monitoring data from all battery cells are recorded synchronously at consistent time points. The time-aligned data is then statistically analyzed hourly, calculating the maximum, average, and minimum internal resistance, voltage, and temperature values ​​for each battery cell during each monitoring period. This generates detailed aggregated operating status data, which is stored simultaneously with the battery's static attribute information.

[0176] In the hospital's battery data processing stage, the system performs category coding and numerical standardization on static attribute information. Brand and model are converted into coded data, while capacity and service life are normalized. The system filters clustering features from the standardized data, extracting attributes that have a significant impact on battery operating status for subsequent grouping. Based on the feature data, the system applies an optimized distance algorithm to group battery cells with similar operating status and consistent attributes into initial battery cell groups. Subsequently, the system checks whether the initial groups maintain a high degree of consistency in key attributes such as brand, model, and capacity. After verification, homogeneous battery cell groups are generated, and the group affiliation of all batteries is recorded.

[0177] For each homogeneous battery cell group, the hospital's power management system analyzes historical changes in internal resistance data and generates a dynamic baseline curve based on the monitoring cycle, forming a real-time reference standard reflecting the battery's health baseline. The system compares the current internal resistance data of each battery cell with the dynamic baseline curve, automatically calculating a standard internal resistance score sequence to identify whether the internal resistance continuously deviates from the baseline range. Through analysis of the standard internal resistance score sequence, the system continuously detects continuous deviations and automatically marks battery cells with long-term internal resistance deviations.

[0178] For the marked battery cells exhibiting persistent deviation, the system further extracts their internal resistance standard score sequence, identifies time segments of persistent internal resistance deviation, and calculates the duration of each deviation segment. Based on the persistent anomaly standards set by the hospital's backup battery management requirements, the system determines whether the deviation time exceeds a threshold, generates a persistent anomaly status judgment result, and records the time node of anomaly confirmation. The system confirms the anomalies of all persistently abnormal battery cells, eliminates short-term fluctuations or the influence of external equipment factors, summarizes them to obtain a preliminary set of abnormal battery cells, and records the generation time.

[0179] The hospital's power management system extracts the internal resistance change rate and voltage fluctuation characteristics of each battery from a preliminary set of abnormal battery cells. Through temperature correlation analysis, it identifies the relationship between battery internal resistance and ambient temperature fluctuations. The system inputs these characteristics into a hospital-customized time-series analysis model to classify and identify whether the abnormal cells are truly degraded or experiencing short-term abnormal interference. The anomaly classification results are processed by a false alarm filtering mechanism, automatically eliminating false alarms caused by non-battery factors such as ambient temperature fluctuations and equipment load changes, thus generating a confirmed list of abnormal battery cells.

[0180] Based on the confirmed list of abnormal battery cells, the hospital's power management system automatically associates the static attribute information of each abnormal battery to generate a detailed list of abnormal batteries, including battery model, capacity, abnormality type, and abnormality time. The system matches maintenance strategies to the detailed list of abnormal batteries using the hospital's battery maintenance knowledge base, determining priorities and maintenance plans for different battery types, such as on-site deep discharge testing, battery replacement, and battery pack readjustment. The system populates the confirmed abnormal battery information and maintenance strategies into the hospital's battery maintenance plan, clearly defining the maintenance task, schedule, responsible person, and operating steps for each abnormal battery. This plan guides the hospital's logistics and power team to efficiently and systematically execute backup battery maintenance tasks, ensuring that the hospital's critical equipment's continuous power supply capability is not affected during mains power failures. The entire process automates and intelligently manages the hospital's backup batteries, reduces the burden of manual analysis, and improves the response speed of battery anomaly detection and the rationality of maintenance decisions.

[0181] In the fintech field, the above-mentioned battery anomaly detection and maintenance methods can effectively ensure the continuity of financial business systems and data security for uninterruptible power supply systems supporting data centers, core computer rooms, or financial trading platforms, and avoid high-risk events such as transaction interruptions and system crashes caused by battery deterioration or failure.

[0182] In the data centers of large financial institutions, multiple sets of lead-acid batteries are deployed for UPS backup systems. These battery units directly determine the emergency power supply capability of the financial business systems in the event of a sudden main power failure. The management system obtains the operating status parameters of all battery units in real time through the environmental monitoring network of the financial data center, including internal resistance, voltage, and temperature data. At the same time, the system synchronously manages the static attribute data of each battery unit, covering information such as brand, model, capacity, and service life. This data is regularly updated and archived by the asset management system of the financial data center.

[0183] During the data acquisition phase, the system first performs data cleaning on all operating status parameters, removing noisy data, abnormal sampled values, and incomplete records to ensure the accuracy of subsequent analysis. The cleaned data is then uniformly aligned to a time base, forming a structured time series to ensure strict time synchronization of monitoring data for each battery cell. Based on the aligned data, the system calculates the maximum, average, and minimum internal resistance, voltage, and temperature values ​​for each battery cell within each hourly cycle, generating stable and reliable aggregated operating status data. This aggregated data, along with static attribute data, is stored on the financial institution's data management platform.

[0184] The system performs categorical attribute coding and numerical attribute standardization on battery static attribute data, converting brand and model into a unified coding format, and standardizing parameters such as capacity and service life into highly comparable numerical expressions. Through feature screening, the system extracts key influencing factors to participate in the subsequent dynamic clustering process. Battery cells in the financial data center are divided into initial battery cell groups based on an optimized distance algorithm. The system verifies the consistency of attributes such as brand, model, and capacity within each group to ensure that the batteries within the group have a high degree of homogeneity, ultimately forming homogeneous battery cell groups and recording the group affiliation information of each battery cell.

[0185] For homogeneous battery cell packs, the system performs dynamic benchmark analysis based on internal resistance data from operating status parameters. By analyzing historical data, the system constructs a dynamic benchmark curve for each pack, reflecting the internal resistance fluctuation range under normal operating conditions. The system automatically calculates a standard internal resistance score sequence, compares the current internal resistance level with the dynamic benchmark curve, identifies battery cells deviating from the normal range in real time, and marks persistent deviations.

[0186] For battery cells exhibiting continuous deviation, the system further acquires a standard internal resistance score sequence, identifies consecutive deviation segments, calculates the duration of each segment, and, based on the continuous anomaly judgment logic, confirms whether the internal resistance deviation possesses long-term sustainability and stability. Based on the continuous anomaly state judgment results and timestamp information, the system identifies anomalous battery cells and compiles them into a preliminary set of anomalous battery cells.

[0187] For each battery in the initial set of abnormal battery cells, the system extracts internal resistance variation characteristics and voltage fluctuation characteristics, and combines these with temperature correlation analysis to comprehensively assess the abnormal battery status. Through a time-series analysis model, the system utilizes multivariate characteristics to determine the authenticity of the anomaly, avoiding interference from short-term fluctuations or environmental factors. A false alarm filtering mechanism further eliminates misjudgments caused by external factors such as temperature changes and load adjustments, resulting in a highly accurate list of confirmed abnormal battery cells.

[0188] Based on the confirmed list of abnormal battery cells, the system correlates the static attribute data of each battery to generate a detailed list, including brand, model, capacity, abnormality type, and time information. Using the maintenance knowledge base of the financial data center, the system performs strategy matching on the detailed list of abnormal batteries to determine priorities, maintenance methods, and operational steps, forming a customized maintenance strategy. Finally, the system combines the strategy matching results with a standardized maintenance plan template to transform the results into a structured battery cell maintenance plan, specifying the maintenance time, operators, and technical requirements for each battery.

[0189] This process enables intelligent anomaly detection and maintenance strategy formulation for UPS backup batteries in the financial technology field, ensuring the stable operation of critical facilities such as financial transaction systems, core data storage devices, and business continuity platforms in the event of main power failure. It reduces the risk of business interruption, data corruption, and economic losses caused by battery performance degradation, and improves the overall operation and maintenance management efficiency and system stability of financial institutions.

[0190] This embodiment generates a detailed battery inventory by combining the confirmed list of abnormal battery cells with static attribute data, providing an accurate basis for subsequent maintenance work. Strategy matching based on a preset maintenance knowledge base ensures that each abnormal battery cell receives a customized maintenance strategy, improving the targeting and efficiency of maintenance. Using standardized templates to populate the maintenance plan ensures the uniformity and operability of maintenance work, thereby reducing the need for manual intervention, improving the automation level of battery management, and ultimately achieving efficient and accurate battery maintenance.

[0191] In one embodiment, a battery cell anomaly detection device is provided, which corresponds one-to-one with the battery cell anomaly detection method described in the above embodiments. (Refer to...) Figure 3 , Figure 3 This is a schematic diagram of the functional modules of a preferred embodiment of the battery cell anomaly detection device of the present invention. The modules include a data acquisition module 10, a cluster analysis module 20, a benchmark analysis module 30, a time series analysis module 40, and a maintenance plan generation module 50. Detailed descriptions of each functional module are as follows:

[0192] The data acquisition module 10 is used to acquire static attribute data and operating status parameters of multiple battery cells;

[0193] Clustering analysis module 20 is used to dynamically cluster the multiple battery cells based on the static attribute data to generate homogeneous battery cell groups;

[0194] The benchmark analysis module 30 is used to perform dynamic benchmark analysis based on the internal resistance data in the operating state parameters within the homogeneous battery cell group, to initially identify abnormal battery cells, and to form a preliminary abnormal battery cell set.

[0195] The time series analysis module 40 is used to perform multivariate time series analysis on each battery cell in the preliminary abnormal battery cell set using a time series analysis model, so as to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells.

[0196] The maintenance plan generation module 50 is used to generate a battery cell maintenance plan table based on the confirmed abnormal battery cell list and the preset maintenance knowledge base.

[0197] In one embodiment, the data acquisition module 10 is specifically used for:

[0198] Collect operating status parameters for each battery cell, including internal resistance, voltage, and temperature;

[0199] Obtain static attribute data for each battery cell, including brand, model, capacity, and service life;

[0200] Perform data cleaning operation on the operating status parameters to generate cleaned operating status parameters;

[0201] Perform a time alignment operation on the cleaned operating status parameters to generate time-aligned operating data;

[0202] Perform statistical aggregation operations on the time-aligned running data to generate running status aggregated data;

[0203] Store the aggregated running status data and the static attribute data.

[0204] In one embodiment, the clustering analysis module 20 is specifically used for:

[0205] Perform categorical attribute encoding and numerical attribute standardization on the static attribute data to generate standardized static attribute data;

[0206] Perform feature selection on the standardized static attribute data to determine clustering feature data;

[0207] Perform distance-optimized grouping on the clustering feature data to generate initial battery cell groups;

[0208] Perform an attribute consistency verification operation on the initial battery cell group to generate a homogeneous battery cell group;

[0209] Record the division results of the homogeneous battery cell group.

[0210] In one embodiment, the benchmark analysis module 30 is specifically used for:

[0211] Based on the internal resistance data in the operating status parameters, a benchmark determination operation is performed on the homogeneous battery cell group to generate dynamic benchmark curve data.

[0212] Perform a standard score determination operation on the internal resistance data of each battery cell to generate a sequence of standard internal resistance scores;

[0213] Perform a persistent deviation detection operation on the internal resistance standard fraction sequence to generate a marker for persistently deviating battery cells;

[0214] An anomaly identification operation is performed on the battery cells whose markers continuously deviate from the target, generating a preliminary set of abnormal battery cells;

[0215] The dynamic reference curve data is updated based on the internal resistance data in the newly acquired operating status parameters.

[0216] In one embodiment, the benchmark analysis module 30 is specifically used for:

[0217] For each marker that continuously deviates from the battery cell, obtain its internal resistance standard score sequence;

[0218] Perform a continuous deviation segment identification operation on the internal resistance standard fraction sequence to generate a continuous deviation segment set;

[0219] Perform a duration determination operation on each consecutive deviation segment to generate a segment duration value;

[0220] Based on the duration value of the segment, an abnormality persistence determination operation is performed to generate a persistent abnormal state determination result, and the generation timestamp of the persistent abnormal state determination result is recorded.

[0221] An anomaly confirmation operation is performed on the continuous abnormal state judgment result to generate a confirmed abnormal battery cell;

[0222] All confirmed abnormal battery cells are aggregated to generate a preliminary abnormal battery cell set, and the generation timestamp of the preliminary abnormal battery cell set is recorded.

[0223] In one embodiment, the timing analysis module 40 is specifically used for:

[0224] For each battery cell in the preliminary abnormal battery cell set, extract the internal resistance change characteristics and voltage fluctuation characteristics from the operating state parameters;

[0225] Based on the internal resistance and temperature data in the operating status parameters, perform a temperature correlation analysis to generate temperature correlation results.

[0226] Using a time-series analysis model, an anomaly classification operation is performed based on the internal resistance change characteristics, the voltage fluctuation characteristics, and the temperature correlation results to generate anomaly classification results.

[0227] Perform a false alarm filtering operation on the anomaly classification results, remove transient interference anomaly classification results, and generate a list of confirmed abnormal battery cells.

[0228] In one embodiment, the maintenance plan generation module 50 is specifically used for:

[0229] For each battery cell in the confirmed abnormal battery cell list, associate its static attribute data to generate a detailed list of abnormal batteries;

[0230] Based on a preset maintenance knowledge base, a maintenance strategy matching operation is performed on the detailed list of abnormal batteries to generate maintenance strategy matching results.

[0231] According to the preset maintenance plan template, the plan filling operation is performed on the maintenance strategy matching result to generate the battery cell maintenance plan.

[0232] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides determination and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used for communication with external user terminals via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a battery cell anomaly detection method on the server side.

[0233] In one embodiment, a computer device is provided, which may be a user terminal, and its internal structure diagram may be as follows: Figure 5As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides determination and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with an external server via a network connection. When executed by the processor, the computer program implements the functions or steps of a battery cell anomaly detection method on the user side.

[0234] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0235] Obtain static attribute data and operating status parameters of multiple battery cells;

[0236] Based on the static attribute data, the multiple battery cells are dynamically clustered to generate homogeneous battery cell groups.

[0237] Within the homogeneous battery cell group, dynamic benchmark analysis is performed based on the internal resistance data in the operating state parameters to initially identify abnormal battery cells and form a preliminary set of abnormal battery cells.

[0238] For each battery cell in the preliminary abnormal battery cell set, a multivariate time series analysis is performed using a time series analysis model to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells.

[0239] Based on the confirmed abnormal battery cell list and the preset maintenance knowledge base, a battery cell maintenance plan is generated.

[0240] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0241] Obtain static attribute data and operating status parameters of multiple battery cells;

[0242] Based on the static attribute data, the multiple battery cells are dynamically clustered to generate homogeneous battery cell groups.

[0243] Within the homogeneous battery cell group, dynamic benchmark analysis is performed based on the internal resistance data in the operating state parameters to initially identify abnormal battery cells and form a preliminary set of abnormal battery cells.

[0244] For each battery cell in the preliminary abnormal battery cell set, a multivariate time series analysis is performed using a time series analysis model to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells.

[0245] Based on the confirmed abnormal battery cell list and the preset maintenance knowledge base, a battery cell maintenance plan is generated.

[0246] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and user side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0247] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0248] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0249] It should be noted that if any software tools or components not belonging to this company appear in the embodiments of this application, they are merely illustrative examples and do not represent actual use. The embodiments described above are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for detecting abnormalities in battery cells, characterized in that, Includes the following steps: Obtain static attribute data and operating status parameters of multiple battery cells; Based on the static attribute data, the multiple battery cells are dynamically clustered to generate homogeneous battery cell groups. Within the homogeneous battery cell group, dynamic benchmark analysis is performed based on the internal resistance data in the operating state parameters to initially identify abnormal battery cells and form a preliminary set of abnormal battery cells. For each battery cell in the preliminary abnormal battery cell set, a multivariate time series analysis is performed using a time series analysis model to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells. Based on the confirmed abnormal battery cell list and the preset maintenance knowledge base, a battery cell maintenance plan is generated.

2. The battery cell anomaly detection method as described in claim 1, characterized in that, Obtain static attribute data and operating status parameters of multiple battery cells, including: Collect operating status parameters for each battery cell, including internal resistance, voltage, and temperature; Obtain static attribute data for each battery cell, including brand, model, capacity, and service life; Perform data cleaning operation on the operating status parameters to generate cleaned operating status parameters; Perform a time alignment operation on the cleaned operating status parameters to generate time-aligned operating data; Perform statistical aggregation operations on the time-aligned running data to generate running status aggregated data; Store the aggregated running status data and the static attribute data.

3. The battery cell anomaly detection method as described in claim 1, characterized in that, Dynamic clustering of the multiple battery cells based on the static attribute data generates homogeneous battery cell groups, including: Perform categorical attribute encoding and numerical attribute standardization on the static attribute data to generate standardized static attribute data; Perform feature selection on the standardized static attribute data to determine clustering feature data; Perform distance-optimized grouping on the clustering feature data to generate initial battery cell groups; Perform an attribute consistency verification operation on the initial battery cell group to generate a homogeneous battery cell group; Record the division results of the homogeneous battery cell group.

4. The battery cell anomaly detection method as described in claim 1, characterized in that, Within the homogeneous battery cell group, dynamic benchmark analysis is performed based on internal resistance data in the operating state parameters to initially identify abnormal battery cells, forming a preliminary set of abnormal battery cells, including: Based on the internal resistance data in the operating status parameters, a benchmark determination operation is performed on the homogeneous battery cell group to generate dynamic benchmark curve data. Perform a standard score determination operation on the internal resistance data of each battery cell to generate a sequence of standard internal resistance scores; Perform a persistent deviation detection operation on the internal resistance standard fraction sequence to generate a marker for persistently deviating battery cells; An anomaly identification operation is performed on the battery cells whose markers continuously deviate from the target, generating a preliminary set of abnormal battery cells; The dynamic reference curve data is updated based on the internal resistance data in the newly acquired operating status parameters.

5. The battery cell anomaly detection method as described in claim 4, characterized in that, An anomaly identification operation is performed on the battery cells whose markers continuously deviate from their positions, generating a preliminary set of abnormal battery cells, including: For each marker that continuously deviates from the battery cell, obtain its internal resistance standard score sequence; Perform a continuous deviation segment identification operation on the internal resistance standard fraction sequence to generate a continuous deviation segment set; Perform a duration determination operation on each consecutive deviation segment to generate a segment duration value; Based on the duration value of the segment, an abnormality persistence determination operation is performed to generate a persistent abnormal state determination result, and the generation timestamp of the persistent abnormal state determination result is recorded. An anomaly confirmation operation is performed on the continuous abnormal state judgment result to generate a confirmed abnormal battery cell; All confirmed abnormal battery cells are aggregated to generate a preliminary abnormal battery cell set, and the generation timestamp of the preliminary abnormal battery cell set is recorded.

6. The battery cell anomaly detection method as described in claim 1, characterized in that, For each battery cell in the preliminary set of abnormal battery cells, a multivariate time-series analysis is performed using a time-series analysis model to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells, including: For each battery cell in the preliminary abnormal battery cell set, extract the internal resistance change characteristics and voltage fluctuation characteristics from the operating state parameters; Based on the internal resistance and temperature data in the operating status parameters, perform a temperature correlation analysis to generate temperature correlation results. Using a time-series analysis model, an anomaly classification operation is performed based on the internal resistance change characteristics, the voltage fluctuation characteristics, and the temperature correlation results to generate anomaly classification results. Perform a false alarm filtering operation on the anomaly classification results, remove transient interference anomaly classification results, and generate a list of confirmed abnormal battery cells.

7. The battery cell anomaly detection method as described in claim 1, characterized in that, Based on the confirmed abnormal battery cell list and the preset maintenance knowledge base, a battery cell maintenance plan is generated, including: For each battery cell in the confirmed abnormal battery cell list, associate its static attribute data to generate a detailed list of abnormal batteries; Based on a preset maintenance knowledge base, a maintenance strategy matching operation is performed on the detailed list of abnormal batteries to generate maintenance strategy matching results. According to the preset maintenance plan template, the plan filling operation is performed on the maintenance strategy matching result to generate the battery cell maintenance plan.

8. A battery cell anomaly detection device, characterized in that, The battery cell anomaly detection device includes: The data acquisition module is used to acquire static attribute data and operating status parameters of multiple battery cells; The clustering analysis module is used to dynamically cluster the multiple battery cells based on the static attribute data to generate homogeneous battery cell groups. The benchmark analysis module is used to perform dynamic benchmark analysis based on the internal resistance data in the operating state parameters within the homogeneous battery cell group, to initially identify abnormal battery cells and form a preliminary set of abnormal battery cells. The time series analysis module is used to perform multivariate time series analysis on each battery cell in the preliminary abnormal battery cell set using a time series analysis model, so as to classify the initially identified abnormal battery cells and generate a list of confirmed abnormal battery cells. The maintenance plan generation module is used to generate a battery cell maintenance plan table based on the confirmed abnormal battery cell list and the preset maintenance knowledge base.

9. A computer device, characterized in that, The computer device includes a memory, a processor, and a battery cell anomaly detection program stored in the memory and executable on the processor. When executed by the processor, the battery cell anomaly detection program implements the steps of the battery cell anomaly detection method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a battery cell anomaly detection program, which, when executed by a processor, implements the steps of the battery cell anomaly detection method as described in any one of claims 1-7.

Images