Battery pack abnormality detection method and system based on average deviation and wavelet analysis

Abstract

The disclosure provides a battery pack anomaly detection method and system based on average deviation and wavelet analysis, relates to the technical field of new energy vehicle power battery pack detection, and comprises the following steps: acquiring voltage data of each battery monomer in a power battery pack in real time and preprocessing; converting the preprocessed voltage data into a time series format and compressing storage by adopting a time partition strategy; calling the stored voltage time series data of each battery monomer, obtaining the voltage average deviation of each battery monomer by using a voltage deviation calculation method; performing wavelet transform on the voltage average deviation of each battery monomer, performing multi-scale decomposition by using a wavelet base function, obtaining high-frequency components of different frequency bands, analyzing the energy distribution of the high-frequency components, calculating statistical characteristics of high-frequency energy, setting an energy threshold of the battery monomer, and determining whether the battery monomer is abnormal by comparing the statistical characteristics with the energy threshold.

Description

Technical Field

[0001] This disclosure relates to the field of power battery pack testing technology for new energy vehicles, specifically to a battery pack anomaly detection method and system based on average deviation and wavelet analysis. Background Technology

[0002] The statements in this section are merely background information relating to this disclosure and do not necessarily constitute prior art.

[0003] The Battery Management System (BMS) in new energy vehicles (such as electric vehicles and hybrid vehicles) manages and controls the power battery pack, directly affecting vehicle safety, range, performance, and battery life. The BMS monitors and tracks the voltage, current, and temperature of each battery cell in real time, enabling management of the battery pack's temperature and control of its charge. Therefore, real-time monitoring of abnormal battery pack voltage is a core aspect of battery management control and a key factor in ensuring safe driving performance.

[0004] Current battery management systems (BMS) primarily rely on set voltage thresholds for alarms, triggering alerts only when the voltage of a single cell exceeds the upper limit or falls below the lower limit. However, this system and method, which mainly relies on fixed voltage thresholds for alarm triggering, cannot identify slow voltage changes or potential abnormal trends, resulting in insufficient ability to detect early signs of failure. The system does not fully mine and utilize historical time-series voltage data, lacks modeling and analysis of long-term voltage change patterns, and cannot conduct predictive maintenance or early intervention. Existing solutions only focus on the instantaneous voltage value of the cell, lacking in-depth analysis and pattern recognition of historical voltage time-series data, and thus cannot provide early warnings of potential battery risks. Summary of the Invention

[0005] To address the aforementioned issues, this disclosure proposes a battery pack anomaly detection method and system based on average deviation and wavelet analysis. The voltage deviation calculation method based on average deviation enables accurate quantification of the consistency level between individual battery cells and the overall battery pack. Simultaneously, wavelet transform is employed to perform multi-scale decomposition of the voltage signal, effectively extracting feature information of the voltage signal in different frequency domains, thus achieving accurate identification of abnormal voltage fluctuations.

[0006] According to some embodiments, the present disclosure adopts the following technical solutions: Battery pack anomaly detection methods based on average deviation and wavelet analysis include: Real-time acquisition and preprocessing of voltage data of each battery cell in the power battery pack; The preprocessed voltage data is converted into a time series format and compressed and stored using a time partitioning strategy. Retrieve the stored voltage time series data of each cell and use the voltage deviation calculation method to obtain the average voltage deviation of each individual cell; Wavelet transform is performed on the average voltage deviation of each individual cell, and multi-scale decomposition is performed using wavelet basis functions to obtain high-frequency components in different frequency bands. The energy distribution of high-frequency components is analyzed, the statistical characteristics of high-frequency energy are calculated, and an energy threshold for the individual cell is set. By comparing the statistical characteristics with the energy threshold, it is determined whether the individual cell is abnormal.

[0007] According to some embodiments, the present disclosure adopts the following technical solutions: A battery pack anomaly detection system based on average deviation and wavelet analysis includes: The data acquisition module is used to acquire and preprocess the voltage data of each battery cell in the power battery pack in real time. The data storage module is used to convert the preprocessed voltage data into a time series format and compress and store it using a time partitioning strategy. The average deviation calculation module is used to retrieve the stored voltage time series data of each cell and use the voltage deviation calculation method to obtain the average voltage deviation of each individual cell. The anomaly detection module is used to perform wavelet transform on the average voltage deviation of each individual cell, use wavelet basis functions for multi-scale decomposition, obtain high-frequency components in different frequency bands, analyze the energy distribution of high-frequency components, calculate the statistical characteristics of high-frequency energy, set the energy threshold of the battery cell, and determine whether the battery cell has an anomaly by comparing the statistical characteristics with the energy threshold.

[0008] According to some embodiments, the present disclosure adopts the following technical solutions: A computer program product includes a computer program that, when executed by a processor, implements the battery pack anomaly detection method based on average deviation and wavelet analysis.

[0009] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the battery pack anomaly detection method based on average deviation and wavelet analysis.

[0010] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform the battery pack anomaly detection method based on average deviation and wavelet analysis.

[0011] Compared with the prior art, the beneficial effects of this disclosure are as follows: This disclosure presents a battery pack anomaly detection method based on average deviation and wavelet analysis. Addressing the voltage consistency assessment problem during power battery pack operation, this method proposes a voltage deviation calculation method based on average deviation. By calculating the difference between the current voltage value of each battery cell and the average voltage of all battery cells at the current moment, it achieves precise quantification of the consistency level between individual battery cells and the overall system. Simultaneously, wavelet transform is used to perform multi-scale decomposition of the voltage signal, effectively extracting feature information of the voltage signal in different frequency domains, enabling accurate identification of abnormal voltage fluctuations. In actual vehicle operation data verification, this method successfully identified multiple voltage outliers. In-depth analysis of the battery cells corresponding to these outliers revealed potential problems within the battery pack, providing important evidence for battery pack health status assessment and fault early warning.

[0012] The battery pack anomaly detection method disclosed herein, based on average deviation and wavelet analysis, retrieves the stored voltage time series data of each cell, uses the voltage deviation calculation method to obtain the average voltage deviation of each individual cell, and obtains the deviation data of each cell at each time point, thereby achieving accurate identification of abnormal voltage fluctuations.

[0013] This disclosure presents a battery pack anomaly detection method based on average deviation and wavelet analysis. This method can quickly and effectively identify abnormal battery cells with voltage deviations from a large amount of driving data from new energy vehicles, and visually display the analysis results through scatter plots and other visualization methods, providing reliable technical support for battery pack maintenance and replacement decisions. The system automatically records and extracts information about these abnormal battery cells, including the battery cell number, the time of anomaly occurrence, and high-frequency energy values, forming a list of abnormal battery cells. Through continuous monitoring and analysis of these abnormal battery cells, potential problems in the battery pack can be identified in a timely manner, providing a basis for battery pack maintenance and replacement decisions. Attached Figure Description

[0014] The accompanying drawings, which form part of this disclosure, are used to provide a further understanding of this disclosure. The illustrative embodiments of this disclosure and their descriptions are used to explain this disclosure and do not constitute an undue limitation of this disclosure.

[0015] Figure 1This is a flowchart of a battery pack anomaly detection method based on average deviation and wavelet analysis according to an embodiment of this disclosure. Detailed Implementation

[0016] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0017] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains.

[0018] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this disclosure. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0019] Example 1 One embodiment of this disclosure provides a battery pack anomaly detection method based on average deviation and wavelet analysis, the steps of which include: Step 1: Acquire and preprocess the voltage data of each battery cell in the power battery pack in real time; Step 2: Convert the preprocessed voltage data into a time series format and compress and store it using a time partitioning strategy; Step 3: Retrieve the stored voltage time series data of each cell and use the voltage deviation calculation method to obtain the average voltage deviation of each individual cell. Step 4: Perform wavelet transform on the average voltage deviation of each individual cell, use wavelet basis functions for multi-scale decomposition, obtain high-frequency components in different frequency bands, analyze the energy distribution of high-frequency components, calculate the statistical characteristics of high-frequency energy, set the energy threshold of the individual cell, and determine whether the individual cell is abnormal by comparing the statistical characteristics with the energy threshold.

[0020] As one embodiment, the battery pack anomaly detection method based on average deviation and wavelet analysis disclosed herein, specifically the voltage deviation calculation method based on average deviation, achieves accurate quantification of the consistency level between individual battery cells and the overall battery pack by calculating the difference between the current voltage value of each individual battery cell and the average voltage value of all individual battery cells at the current moment. Simultaneously, wavelet transform is used to perform multi-scale decomposition of the voltage signal, effectively extracting the feature information of the voltage signal in different frequency domains, thus achieving accurate identification of abnormal voltage fluctuations. The specific implementation process of the method is as follows: Step 1: Acquire and preprocess the voltage data of each battery cell in the power battery pack in real time; Specifically, the voltage data of each battery cell in the power battery pack is acquired in real time. This includes: during the data acquisition process, the voltage data of each battery cell in the power battery pack is acquired in real time through the vehicle BMS system. The data is acquired according to the set sampling frequency. The acquired data includes the battery cell number, voltage value and acquisition timestamp information. Then, it is transmitted to the vehicle terminal through the CAN bus and uploaded to the cloud server in real time.

[0021] Next, the data is stored. Before storing the data, the voltage data is preprocessed, including data cleaning, outlier removal, and timestamp alignment.

[0022] Furthermore, the preprocessed voltage data is stored in a time-series format, with each data point including a timestamp, battery cell number, and voltage value. The database employs a time-partitioned storage strategy, sharding and partitioning tables according to time ranges, supporting fast querying and statistical analysis by time range. Simultaneously, the system periodically compresses and stores historical data to optimize storage space utilization efficiency.

[0023] Step 2: Retrieve the stored voltage time series data of each cell and use the voltage deviation calculation method to obtain the average voltage deviation of each individual cell. Specifically, the process of calculating the average voltage deviation of a single battery cell includes: Suppose at any time t have N Each battery cell has the following voltages:

[0024] Then the average voltage of all cells at that moment is:

[0025] Then the battery cell i The average deviation is:

[0026] Step 3: Perform wavelet transform on the average voltage deviation of each individual cell, use wavelet basis functions for multi-scale decomposition, obtain high-frequency components in different frequency bands, analyze the energy distribution of high-frequency components, calculate the statistical characteristics of high-frequency energy, set the energy threshold of the battery cell, and determine whether the battery cell is abnormal by comparing the statistical characteristics with the energy threshold.

[0027] Specifically, firstly, a wavelet transform is performed on the average voltage deviation of each individual cell. Then, pywt.wavedec(series, wavelet) is called to decompose the input time series into a set of different frequency components: Approximation coefficients: Low-frequency components, representing overall trends or long-term changes.

[0028] Detail Coefficients: High-frequency components, representing short-term fluctuations at different levels. The default is db1 (Daubechies 1 wavelet).

[0029] low_freq_component = coeffs[0]: Extracts the lowest frequency approximation coefficient (overall trend).

[0030] `selected_high_freq_component = coeffs[detail_level]`: Extracts high-frequency components from a specific layer. During the decomposition process, by setting an appropriate number of decomposition layers and a target time scale, it is ensured that high-frequency and low-frequency feature information in the voltage signal can be effectively extracted.

[0031] The process of calculating the number of layers is as follows: With each additional layer in wavelet decomposition, the temporal resolution of the signal is halved. Therefore, the following formula can be used to calculate the number of layers:

[0032] As one example, if the sampling interval is 10 seconds (i.e., one voltage data point is recorded every 10 seconds), then: 12 hours (43200 seconds) =

[0033] 24 hours (86400 seconds) =

[0034] Therefore, in wavelet decomposition, layers 12 and 13 will contain periodic components of approximately 12 hours and 24 hours, respectively. Choosing the detail coefficients for layer 12 or 13 can help analyze the 12- to 24-hour periodic fluctuations in the data.

[0035] Furthermore, the calculation of low-frequency and high-frequency energies includes: Low-frequency energy:

[0036] in, For the first Approximation coefficients of the layer For the first Each coefficient is taken from the wavelet decomposition result (the lowest frequency coefficient). The square of each coefficient represents the energy contribution of that coefficient. The sum of all the squared results is used to obtain the total low-frequency energy.

[0037] High-frequency energy (layer 1):

[0038] Among them, the detail coefficient of the first layer represents the high-frequency (rapid fluctuation) part.

[0039] Furthermore, for the extracted high-frequency energy, its energy value is calculated using the following formula:

[0040] Where c is a high-frequency coefficient.

[0041] By analyzing the energy distribution of high-frequency components, abrupt changes and abnormal fluctuations in the voltage signal can be identified, including: calculating the statistical characteristics of high-frequency energy, including the mean and standard deviation. When the low-frequency energy and high-frequency energy values ​​of a certain battery cell exceed three times the standard deviation, the battery cell is determined to be abnormal.

[0042] For each cell i, determine whether one of the following conditions is met:

[0043] Add the battery cells that meet the above conditions to set S:

[0044] As an example, wavelet energy analysis was performed on cell voltage_42, yielding a low-frequency energy of 18.67 and a high-frequency energy of 2.67. The relatively high low-frequency energy (18.67) indicates that the cell voltage exhibits a significant trend or slow variation over a long time scale, reflecting voltage drift, load accumulation effects, or overall performance degradation during operation. High low-frequency energy typically signifies the presence of a strong slow-varying structure or steady-state offset in the system.

[0045] The high-frequency energy (2.67) is much higher than other cells, indicating that the cell experiences severe rapid fluctuations or disturbances within a long period of 12–24 hours. This may be due to the following factors: frequent or abnormal charge and discharge operations; frequent fluctuations in external load causing voltage instability; internal physical or chemical abnormalities, such as uneven internal resistance, poor contact, or abnormal electrolyte distribution.

[0046] Based on statistical analysis, when the high-frequency or low-frequency energy of a certain battery cell exceeds three times the standard deviation (3σ) of the average value of similar battery cells, it can be preliminarily determined that it is in an atypical state or has potential abnormal operation, and it is recommended to further track or issue an early warning for the battery cell.

[0047] For each cell i, determine whether one of the following conditions is met:

[0048] Add the battery cells that meet the above conditions to set S:

[0049] This disclosure focuses on "rapid disturbances" in high-frequency energy and "long-term trends" in low-frequency energy; combining the two provides a comprehensive characterization of the cell's operating status. High-frequency energy typically reflects rapid fluctuations in the signal, while a significantly higher long-cycle high-frequency energy level than other cells may indicate that the voltage of voltage_42 is unstable over a longer period, exhibiting abnormal fluctuations. This instability could be caused by factors such as cell degradation, capacity imbalance, or increased internal impedance, all of which can lead to larger voltage fluctuations during normal charging and discharging. The significantly higher long-cycle high-frequency energy level of voltage_42 compared to other cells suggests significant voltage fluctuations within a 12-24 hour period. These fluctuations may originate from internal instabilities within the cell, external environmental interference, or even be early signs of degradation. Based on the test results, further health checks and monitoring of voltage_42 are recommended to ensure it does not cause greater performance problems in future use.

[0050] Example 2 One embodiment of this disclosure provides a battery pack anomaly detection system based on average deviation and wavelet analysis, comprising: The data acquisition module is used to acquire and preprocess the voltage data of each battery cell in the power battery pack in real time. The data storage module is used to convert the preprocessed voltage data into a time series format and compress and store it using a time partitioning strategy. The average deviation calculation module is used to retrieve the stored voltage time series data of each cell and use the voltage deviation calculation method to obtain the average voltage deviation of each individual cell. The anomaly detection module is used to perform wavelet transform on the average voltage deviation of each individual cell, use wavelet basis functions for multi-scale decomposition, obtain high-frequency components in different frequency bands, analyze the energy distribution of high-frequency components, calculate the statistical characteristics of high-frequency energy, set the energy threshold of the battery cell, and determine whether the battery cell has an anomaly by comparing the statistical characteristics with the energy threshold.

[0051] Example 3 One embodiment of this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the battery pack anomaly detection method based on average deviation and wavelet analysis.

[0052] Example 4 One embodiment of this disclosure provides a non-transitory computer-readable storage medium for storing computer instructions. When these computer instructions are executed by a processor, they implement the battery pack anomaly detection method based on average deviation and wavelet analysis.

[0053] Example 5 One embodiment of this disclosure provides an electronic device, including a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform the battery pack anomaly detection method based on average deviation and wavelet analysis.

[0054] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0056] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this disclosure are still within the scope of protection of this disclosure.

Claims

1. A battery pack anomaly detection method based on average deviation and wavelet analysis, characterized in that, include: Real-time acquisition and preprocessing of voltage data of each battery cell in the power battery pack; The preprocessed voltage data is converted into a time series format and compressed and stored using a time partitioning strategy. Retrieve the stored voltage time series data of each cell and use the voltage deviation calculation method to obtain the average voltage deviation of each individual cell; The process of calculating the average voltage deviation of a single battery cell includes: assuming at any given time... t have N Each battery cell has a voltage of [number] respectively. Then, at that moment, the average voltage of all battery cells is: Then the battery cell i The average deviation is: ; Wavelet transform is performed on the average voltage deviation of each individual cell, and multi-scale decomposition is performed using wavelet basis functions to obtain high-frequency components in different frequency bands. The energy distribution of high-frequency components is analyzed, the statistical characteristics of high-frequency energy are calculated, and the energy threshold of the individual cell is set. By comparing the statistical characteristics with the energy threshold, it is determined whether the individual cell is abnormal. For the extracted high-frequency energy, its energy value is calculated using the following formula: Where c is the high-frequency coefficient; by analyzing the energy distribution of high-frequency components, abrupt changes and abnormal fluctuations in the voltage signal are identified. At the same time, the statistical characteristics of high-frequency energy are calculated, including the mean and standard deviation. When the low-frequency energy or high-frequency energy value of a certain battery cell exceeds three times the standard deviation of the corresponding mean, the battery cell is judged to be abnormal.

2. The battery pack anomaly detection method based on average deviation and wavelet analysis as described in claim 1, characterized in that, Real-time acquisition of voltage data of each battery cell in the power battery pack includes: During the data acquisition process, the voltage data of each battery cell in the power battery pack is collected in real time through the vehicle BMS system. The data is collected according to the set sampling frequency. The collected data includes the battery cell number, voltage value and collection timestamp information. Then, it is transmitted to the vehicle terminal through the CAN bus and uploaded to the cloud server in real time.

3. The battery pack anomaly detection method based on average deviation and wavelet analysis as described in claim 1, characterized in that, Before data storage, the voltage data is preprocessed, including data cleaning, outlier removal, and timestamp alignment. The preprocessed voltage data is then stored in a time-series format, with each data point containing a timestamp, battery cell number, and voltage value field. The database uses a time-partitioned storage strategy, supporting fast querying and statistical analysis by time range.

4. The battery pack anomaly detection method based on average deviation and wavelet analysis as described in claim 1, characterized in that, Wavelet transform is performed on the average voltage deviation of each individual cell, and multi-scale decomposition is performed using the db4 order wavelet basis function. The average voltage deviation is decomposed into high-frequency and low-frequency components in different frequency bands. During the decomposition process, by setting the number of decomposition layers and the target time scale, it is ensured that the high-frequency and low-frequency feature information in the voltage signal can be effectively extracted. For each additional layer of wavelet decomposition, the time resolution of the signal is halved.

5. A battery pack anomaly detection system based on average deviation and wavelet analysis, characterized in that, The battery pack anomaly detection method based on average deviation and wavelet analysis as described in any one of claims 1-4 includes: The data acquisition module is used to acquire and preprocess the voltage data of each battery cell in the power battery pack in real time. The data storage module is used to convert the preprocessed voltage data into a time series format and compress and store it using a time partitioning strategy. The average deviation calculation module is used to retrieve the stored voltage time series data of each cell and use the voltage deviation calculation method to obtain the average voltage deviation of each individual cell. The anomaly detection module is used to perform wavelet transform on the average voltage deviation of each individual cell, use wavelet basis functions for multi-scale decomposition, obtain high-frequency components in different frequency bands, analyze the energy distribution of high-frequency components, calculate the statistical characteristics of high-frequency energy, set the energy threshold of the battery cell, and determine whether the battery cell has an anomaly by comparing the statistical characteristics with the energy threshold.

6. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the battery pack anomaly detection method based on average deviation and wavelet analysis as described in any one of claims 1-4.

7. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the battery pack anomaly detection method based on average deviation and wavelet analysis as described in any one of claims 1-4.

8. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform the battery pack anomaly detection method based on average deviation and wavelet analysis as described in any one of claims 1-4.

Images