WO2025260641A1 - 电池电芯故障诊断方法及装置 - Google Patents

电池电芯故障诊断方法及装置

Info

Publication number
WO2025260641A1
WO2025260641A1 PCT/CN2024/138428 CN2024138428W WO2025260641A1 WO 2025260641 A1 WO2025260641 A1 WO 2025260641A1 CN 2024138428 W CN2024138428 W CN 2024138428W WO 2025260641 A1 WO2025260641 A1 WO 2025260641A1
Authority
WO
WIPO (PCT)
Prior art keywords
cell
slope
scale factor
euclidean distance
sum
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/CN2024/138428
Other languages
English (en)
French (fr)
Inventor
苟斌
戴欣
张薇
孟玉航
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Dongfeng Motor Group Co Ltd
Original Assignee
Dongfeng Motor Group Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Dongfeng Motor Group Co Ltd filed Critical Dongfeng Motor Group Co Ltd
Publication of WO2025260641A1 publication Critical patent/WO2025260641A1/zh
Pending legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/385—Arrangements for measuring battery or accumulator variables
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering

Definitions

  • This invention relates to the field of battery technology, and in particular to a method and apparatus for diagnosing battery cell faults.
  • a battery typically consists of multiple cells.
  • a cell malfunction can lead to abnormal charging and discharging, or even fire. To ensure battery safety, it is necessary to diagnose the battery cells.
  • This invention solves the technical problem of how to diagnose battery cell faults by providing a method and apparatus for diagnosing battery cell faults.
  • a method for diagnosing battery cell faults comprising:
  • the sum of the absolute values of MN of the battery cell under test and the Euclidean distance are calculated based on the multiple charging voltages mentioned above.
  • the slope of the sample entropy-scale factor curve of the battery cell under test is calculated based on multiple sampling times.
  • the first Hamming proximity is calculated based on the sum of the absolute values of MN of the cell under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster center of the normal cell.
  • the second Hamming proximity is calculated based on the sum of the absolute values of MN of the cells under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster center of the faulty cells.
  • the cell under test is determined to be faulty.
  • a battery cell fault diagnosis device comprising:
  • the acquisition module is used to acquire the charging voltage of multiple battery cells collected at multiple sampling times;
  • the calculation module is used to calculate the sum of the absolute values of MN of the battery cell under test and the Euclidean distance based on the multiple charging voltages.
  • the slope of the sample entropy-scale factor curve of the battery cell under test is calculated based on multiple sampling times.
  • the acquisition module is also used to acquire preset normal cell cluster centers and faulty cell cluster centers;
  • the calculation module is also used to calculate the first Hamming proximity based on the sum of the absolute values of MN of the cell under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster center of the normal cell.
  • the second Hamming proximity is calculated based on the sum of the absolute values of MN of the cells under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster center of the faulty cells.
  • the judgment module is used to determine that the battery cell under test is faulty if the second Hamming proximity is greater than the first Hamming proximity.
  • a computer device includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of a method for diagnosing faults in any battery cell.
  • a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a method for diagnosing faults in any battery cell.
  • This invention calculates the sum of absolute values of MN and Euclidean distance of the tested cell based on the charging voltage of the battery cell. It also calculates the slope of the sample entropy-scale factor curve of the tested cell based on the sampling time of the charging voltage. Furthermore, it calculates the first Hamming proximity of the sum of absolute values of MN, Euclidean distance, and slope of the sample entropy-scale factor curve of the tested cell to the cluster centers of normal cells. Finally, it calculates the second Hamming proximity of the sum of absolute values of MN, Euclidean distance, and slope of the sample entropy-scale factor curve of the tested cell to the cluster centers of faulty cells. If the second Hamming proximity is greater than the first Hamming proximity, the tested cell is judged to be faulty. This invention provides a comprehensive diagnosis of battery cell faults from three dimensions.
  • Figure 1 is a graph showing the change of cell charging voltage over time in an embodiment of the present invention
  • Figure 2 is a graph showing the change of the MN value of the battery cell with the sampling time in an embodiment of the present invention
  • Figure 3 is a schematic diagram of the sum of the absolute values of MN of the battery cells in an embodiment of the present invention.
  • Figure 4 is a sample entropy-scale factor curve of the battery cell in an embodiment of the present invention.
  • Figure 5 is a schematic diagram of the slope of the sample entropy-scale factor curve in an embodiment of the present invention.
  • Figure 6 is a schematic diagram of the Euclidean distance of the battery cells in an embodiment of the present invention.
  • FIG. 7 is a flowchart of the battery cell fault diagnosis method in an embodiment of the present invention.
  • Figure 8 is a schematic diagram of the battery cell fault diagnosis device in an embodiment of the present invention.
  • the embodiments of the present invention provide a method and apparatus for diagnosing battery cell faults, thereby solving the technical problem of how to diagnose battery cell faults.
  • the embodiments of the present invention obtained the curves of the charging voltage of the 6 battery cells changing over time as shown in Figure 1 through experiments.
  • the curve at the bottom of Figure 1 is cell #5. It can be seen that the charging voltage of cell #5 is too low, and it is a faulty cell.
  • the experiment yielded the curves showing the change of MN values over time for the six cells, as shown in Figure 2.
  • the top curve in Figure 2 represents the MN value of the faulty cell #5.
  • the sum of the absolute MN values of the six cells was calculated.
  • Figure 3 shows that the sum of the absolute MN values of the faulty cell #5 is significantly greater than that of the other cells, indicating that the sum of the absolute MN values can reflect the difference between the faulty cell and the normal cells; in other words, the sum of the absolute MN values can indicate cell faults.
  • the experiment yielded sample entropy-scale factor curves for six battery cells, as shown in Figure 4.
  • the bottom curve in Figure 4 represents the sample entropy-scale factor curve for the faulty cell #5.
  • the slope of each sample entropy-scale factor curve is shown in Figure 5. It can be seen that the slope of the sample entropy-scale factor curve for the faulty cell #5 is significantly lower than that of the other cells, indicating that the slope of the sample entropy-scale factor curve can reflect the cell fault.
  • the experiment yielded the Euclidean distances of the six cells shown in Figure 6. It can be seen that the Euclidean distance of the faulty cell #5 is significantly greater than that of the other cells.
  • the Euclidean distance between the surface cell charging voltage curve and the average voltage curve can also reflect the cell fault.
  • this invention proposes a battery cell fault diagnosis method, as shown in Figure 7, including:
  • Step S1 Obtain the charging voltage of multiple battery cells collected at multiple sampling times
  • Step S2 Calculate the sum of the absolute values of MN of the cell under test and the Euclidean distance based on multiple charging voltages;
  • Step S3 Calculate the slope of the sample entropy-scale factor curve of the cell under test based on multiple sampling times;
  • Step S4 Obtain the preset cluster centers for normal battery cells and faulty battery cells
  • Step S5 Calculate the first Hamming proximity based on the sum of the absolute values of MN of the cells under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster centers of normal cells.
  • Step S6 Calculate the second Hamming proximity based on the sum of the absolute values of MN of the cells under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster center of the faulty cells.
  • Step S7 If the second Hamming proximity is greater than the first Hamming proximity, then the cell under test is determined to be faulty.
  • step S1 assuming that the charging voltage of multiple battery cells is collected every 1 second for a total of 20 seconds, the multiple sampling times are 1, 2, 3, ..., 19, 20.
  • the sum of the absolute values of MN based on multiple charging voltages includes: determining the maximum value among the multiple charging voltages collected at the sampling time; calculating the average value of the multiple charging voltages collected at that sampling time; calculating the MN value corresponding to that sampling time based on the charging voltage, maximum value, and average value of the cell under test collected at that sampling time; and calculating the sum of the absolute values of the multiple MN values corresponding to multiple sampling times to obtain the sum of the absolute values of MN.
  • MN MN is the MN value corresponding to a certain sampling time
  • U the charging voltage of the battery cell under test collected at that sampling time. This represents the average charging voltage of all battery cells collected at that sampling time. This is the maximum value among all cell charging voltages collected at that sampling time.
  • Each sampling time corresponds to a sum of absolute values of MN, and we can obtain the sum of all absolute values of MN for all sampling times. Taking the absolute value of MN prevents the cell's
  • the Euclidean distance is calculated based on multiple charging voltages, including: calculating the average value of multiple charging voltages collected at the sampling time; calculating the difference between the charging voltage of the cell under test collected at that sampling time and the average value; and calculating the Euclidean distance based on the multiple differences corresponding to multiple sampling times.
  • n is the sequence number of the sampling time, and n is the number of sampling times (e.g., 20).
  • Step S3 includes: downsampling multiple sampling times according to multiple different scale factors to obtain multiple sampling times under each scale factor; calculating the sample entropy corresponding to the scale factor based on the multiple sampling times under the scale factor; fitting multiple scale factors and the corresponding multiple sample entropies to obtain the sample entropy-scale factor curve; and determining the slope of the sample entropy-scale factor curve.
  • the normal cell cluster centers include the normal cell MN cluster center, the normal cell Euclidean distance cluster center, and the normal cell slope cluster center;
  • the faulty cell cluster centers include the faulty cell MN cluster center, the faulty cell Euclidean distance cluster center, and the faulty cell slope cluster center.
  • Step S4 specifically includes: obtaining the sum of the absolute values of MN, Euclidean distance, and slope of the sample entropy-scale factor curve for each training sample cell in multiple training sample cells; normalizing the sum of the absolute values of MN, the Euclidean distance, and the slope of the sample entropy-scale factor curve; taking the absolute value of the normalized slope of the sample entropy-scale factor curve, and assigning weights to the absolute values of the sum of the absolute values of MN, the Euclidean distance, and the slope of the sample entropy-scale factor curve; performing k-means clustering on the weighted sum of the absolute values of MN to obtain the MN cluster centers of normal cells and the MN cluster centers of faulty cells; performing k-means clustering on the weighted Euclidean distances to obtain the Euclidean distance cluster centers of normal cells and the Euclidean distance cluster centers of faulty cells; and performing k-means clustering on the absolute values of the weighted slopes of
  • steps S1-S3 are performed by replacing the test cell with each training sample cell to obtain the sum of the absolute values of MN, the Euclidean distance, and the slope of the sample entropy-scale factor curve for each training sample cell.
  • Normalization can be performed using the Z-score normalization method. Taking the absolute value of the slope of the normalized sample entropy-scale factor curve is for the convenience of horizontal comparison. The slope of the sample entropy-scale factor curve is of high importance, and weights of the sum of the absolute values of MN, the Euclidean distance, and the absolute value of the slope of the sample entropy-scale factor curve can be assigned in a weight ratio of 1:1:2.
  • the k-means clustering process is as follows: First, randomly select two sums of absolute values of MN as initial centroids; second, assign each sum of absolute values of MN to the nearest initial centroid, forming two clusters; third, calculate the average of the sums of absolute values of MN in the cluster containing the first initial centroid and use it as the new centroid of that cluster, and calculate the average of the sums of absolute values of MN in the cluster containing the second initial centroid and use it as the new centroid of that cluster; fourth, repeat steps two and three until the centroids of the two clusters no longer change.
  • the centroids of the two clusters are the MN cluster centers of normal cells and the MN cluster centers of faulty cells.
  • the k-means clustering process using Euclidean distance and the slope of the sample entropy-scale factor curve is the same as that for the sum of absolute values of MN.
  • the objective function of k-means clustering is SSE.
  • the values are, in order, the sum of the absolute values of MN of the cell under test, the Euclidean distance, and the slope of the sample entropy-scale factor curve.
  • the cluster centers are, in order: MN cluster center for normal cells, Euclidean distance cluster center for normal cells, and slope cluster center for normal cells.
  • the values are, in order, the sum of the absolute values of MN of the cell under test, the Euclidean distance, and the slope of the sample entropy-scale factor curve.
  • the cluster centers are, in order: MN cluster center of faulty cells, Euclidean distance cluster center of faulty cells, and slope cluster center of faulty cells.
  • the battery cell under test is determined to be normal if the second Hamming proximity of the battery cell under test is not greater than the first Hamming proximity, then the battery cell under test is determined to be normal.
  • the battery cell fault diagnosis method of this invention calculates the sum of absolute values of MN and Euclidean distance of the cell under test based on the charging voltage of the battery cell, calculates the slope of the sample entropy-scale factor curve of the cell under test based on the sampling time of the charging voltage, calculates the first Hamming proximity of the sum of absolute values of MN, Euclidean distance, and slope of the sample entropy-scale factor curve of the cell under test with the cluster center of normal cells, and calculates the second Hamming proximity of the sum of absolute values of MN, Euclidean distance, and slope of the sample entropy-scale factor curve of the cell under test with the cluster center of faulty cells. If the second Hamming proximity is greater than the first Hamming proximity, the cell under test is judged to be faulty. The battery cell fault is diagnosed comprehensively from three dimensions.
  • this embodiment of the invention also provides a battery cell fault diagnosis device, comprising:
  • the acquisition module is used to acquire the charging voltage of multiple battery cells collected at multiple sampling times;
  • the calculation module is used to calculate the sum of the absolute values of MN of the battery cell under test and the Euclidean distance based on multiple charging voltages.
  • the slope of the sample entropy-scale factor curve of the battery cell under test is calculated based on multiple sampling times.
  • the acquisition module is also used to acquire preset cluster centers for normal battery cells and cluster centers for faulty battery cells;
  • the calculation module is also used to calculate the first Hamming proximity based on the sum of the absolute values of MN of the cells under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster centers of normal cells.
  • the second Hamming proximity is calculated based on the sum of the absolute values of MN of the cells under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster center of the faulty cells.
  • the judgment module is used to determine that the cell under test is faulty if the second Hamming proximity is greater than the first Hamming proximity.
  • the calculation module can also be used to: determine the maximum value among multiple charging voltages collected at the sampling time; calculate the average value of multiple charging voltages collected at the sampling time; calculate the MN value corresponding to the sampling time based on the charging voltage, maximum value, and average value of the cell under test collected at the sampling time; and calculate the sum of the absolute values of multiple MN values corresponding to multiple sampling times to obtain the sum of the absolute values of MN.
  • the calculation module can also be used to: calculate the average value of multiple charging voltages collected at the sampling time; calculate the difference between the charging voltage of the cell under test collected at the sampling time and the average value; and calculate the Euclidean distance based on the multiple differences corresponding to multiple sampling times.
  • the calculation module can also be used to: downsample multiple sampling times according to multiple different scale factors to obtain multiple sampling times under each scale factor; calculate the sample entropy corresponding to the scale factor based on the multiple sampling times under the scale factor; fit multiple scale factors and the corresponding multiple sample entropies to obtain the sample entropy-scale factor curve; and determine the slope of the sample entropy-scale factor curve.
  • the normal cell cluster centers include the normal cell MN cluster centers, the normal cell Euclidean distance cluster centers, and the normal cell slope cluster centers;
  • the faulty cell cluster centers include the faulty cell MN cluster centers, the faulty cell Euclidean distance cluster centers, and the faulty cell slope cluster centers.
  • the acquisition module can also be used to: acquire the sum of the absolute values of MN, Euclidean distance, and slope of the sample entropy-scale factor curve for each training sample cell in multiple training sample cells; normalize the sum of the absolute values of MN, the Euclidean distance, and the slope of the sample entropy-scale factor curve; take the absolute value of the normalized slope of the sample entropy-scale factor curve, and assign weights to the absolute values of the sum of the absolute values of MN, Euclidean distance, and slope of the sample entropy-scale factor curve; perform k-means clustering on the weighted sum of the absolute values of MN to obtain the MN cluster centers of normal cells and the MN cluster centers of faulty cells; perform k-means clustering on the weighted Euclidean distance to obtain the Euclidean distance cluster centers of normal cells and the Euclidean distance cluster centers of faulty cells; and perform k-means clustering on the absolute values of the weighted slopes of the sample
  • the calculation module calculates the first Hamming proximity based on the sum of the absolute values of MN of the cells under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster centers of normal cells. This may include:
  • N1 represents the first Hamming proximity score.
  • the values are, in order, the sum of the absolute values of MN of the cell under test, the Euclidean distance, and the slope of the sample entropy-scale factor curve.
  • the cluster centers are, in order: MN cluster center for normal cells, Euclidean distance cluster center for normal cells, and slope cluster center for normal cells.
  • the calculation module calculates the second Hamming proximity based on the sum of the absolute values of MN of the cells under test, the Euclidean distance, the slope of the sample entropy-scale factor curve, and the cluster centers of the faulty cells. This may include:
  • N2 is the second Hamming proximity score.
  • the values are, in order, the sum of the absolute values of MN of the cell under test, the Euclidean distance, and the slope of the sample entropy-scale factor curve.
  • the cluster centers are, in order: MN cluster center of faulty cells, Euclidean distance cluster center of faulty cells, and slope cluster center of faulty cells.
  • this embodiment of the invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any of the battery cell fault diagnosis methods described above.
  • the bus architecture (represented by a bus) can include any number of interconnected buses and bridges, linking various circuits including one or more processors (represented by a processor) and memory (represented by memory).
  • the bus can also link various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein.
  • the bus interface provides an interface between the bus and receivers and transmitters. Receivers and transmitters can be the same element, a transceiver, providing a unit for communicating with various other devices over a transmission medium.
  • the processor is responsible for managing the bus and general processing, while memory can be used to store data used by the processor during operation.
  • the computer device described in this embodiment of the invention is the computer device used to implement the battery cell fault diagnosis method in this embodiment, those skilled in the art can understand the specific implementation methods and various variations of the computer device in this embodiment based on the battery cell fault diagnosis method described in this embodiment. Therefore, how the computer device implements the method in this embodiment will not be described in detail here. Any computer device used by those skilled in the art to implement the battery cell fault diagnosis method in this embodiment falls within the scope of protection of this invention.
  • the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the aforementioned battery cell fault diagnosis methods.
  • embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
  • a computer-usable storage media including, but not limited to, disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and/or one or more block diagrams.
  • These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and/or one or more block diagrams.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Probability & Statistics with Applications (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • Algebra (AREA)
  • Operations Research (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Secondary Cells (AREA)

Abstract

本发明公开了电池电芯故障诊断方法及装置，涉及电池技术领域。本发明根据电池电芯的充电电压计算待测电芯的MN绝对值之和以及欧氏距离，根据充电电压采样时刻计算待测电芯的样本熵-尺度因子曲线斜率，计算待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率与正常电芯聚类中心的第一汉明贴近度，计算待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率与故障电芯聚类中心的第二汉明贴近度，若第二汉明贴近度大于第一汉明贴近度则判断待测电芯故障。

Description

电池电芯故障诊断方法及装置
相关申请的交叉引用
本申请实施例基于申请号为202410809826.6、申请日为2024年06月21日的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此引入本申请实施例作为参考。
技术领域
本发明涉及电池技术领域，尤其涉及电池电芯故障诊断方法及装置。
 背景技术
电池一般包括多个电芯，电芯出现故障会导致电池充放电异常，甚至起火燃烧，为了确保电池安全，需要对电池电芯进行故障诊断。
 发明内容
本发明通过提供电池电芯故障诊断方法及装置，解决了如何诊断电池电芯故障的技术问题。
一方面，本发明提供如下技术方案：
一种电池电芯故障诊断方法，包括：
获取多个采样时刻采集的电池多个电芯的充电电压；
根据多个所述充电电压计算待测电芯的MN绝对值之和以及欧氏距离；
根据多个所述采样时刻计算所述待测电芯的样本熵-尺度因子曲线斜率；
获取预设的正常电芯聚类中心和故障电芯聚类中心；
根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述正常电芯聚类中心计算第一汉明贴近度；
根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述故障电芯聚类中心计算第二汉明贴近度；
若所述第二汉明贴近度大于所述第一汉明贴近度，则判断所述待测电芯故障。
另一方面，本发明还提供如下技术方案：
一种电池电芯故障诊断装置，包括：
获取模块，用于获取多个采样时刻采集的电池多个电芯的充电电压；
计算模块，用于根据多个所述充电电压计算待测电芯的MN绝对值之和以及欧氏距离；
根据多个所述采样时刻计算所述待测电芯的样本熵-尺度因子曲线斜率；
所述获取模块，还用于获取预设的正常电芯聚类中心和故障电芯聚类中心；
所述计算模块，还用于根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述正常电芯聚类中心计算第一汉明贴近度；
根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述故障电芯聚类中心计算第二汉明贴近度；
判断模块，用于若所述第二汉明贴近度大于所述第一汉明贴近度，则判断所述待测电芯故障。
另一方面，本发明还提供如下技术方案：
一种计算机设备，包括存储器、处理器及存储在存储器上的计算机程序，所述处理器执行所述计算机程序以实现任一电池电芯故障诊断方法的步骤。
另一方面，本发明还提供如下技术方案：
一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现任一电池电芯故障诊断方法的步骤。
本发明提供的一个或多个技术方案，至少具有如下技术效果或优点：
本发明根据电池电芯的充电电压计算待测电芯的MN绝对值之和以及欧氏距离，根据充电电压采样时刻计算待测电芯的样本熵-尺度因子曲线斜率，计算待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率与正常电芯聚类中心的第一汉明贴近度，计算待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率与故障电芯聚类中心的第二汉明贴近度，若第二汉明贴近度大于第一汉明贴近度则判断待测电芯故障，从三个维度综合诊断了电池电芯故障。
 附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本发明实施例中电芯充电电压随时间变化的曲线图；
图2为本发明实施例中电芯的MN值随采样时刻变化的曲线图；
图3为本发明实施例中电芯的MN绝对值之和的示意图；
图4为本发明实施例中电芯的样本熵-尺度因子曲线图；
图5为本发明实施例中样本熵-尺度因子曲线斜率的示意图；
图6为本发明实施例中电芯的欧式距离的示意图；
图7为本发明实施例中电池电芯故障诊断方法的流程图；
图8为本发明实施例中电池电芯故障诊断装置的示意图。
 本发明的实施方式
本发明实施例通过提供电池电芯故障诊断方法及装置，解决了如何诊断电池电芯故障的技术问题。
为了更好的理解本发明的技术方案，下面将结合说明书附图以及具体的实施方式对本发明的技术方案进行详细的说明。
本发明实施例经过实验得到了如图1所示的电池6个电芯的充电电压随时间变化的曲线，图1最下方的曲线为5#电芯，可以看出，5#电芯的充电电压偏低，为故障电芯。
经过实验得到如图2所示的6个电芯的MN值随时间变化的曲线，图2最上方的曲线为5#故障电芯的MN值曲线。根据图2计算得到了如图3所示的6个电芯的MN绝对值之和。从图3可以看出，5#故障电芯的MN绝对值之和明显大于其它电芯，表明MN绝对值之和可以体现出故障电芯相对于正常电芯的差异，也就是MN绝对值之和可以体现电芯故障。
经过实验得到如图4所示的6个电芯的样本熵-尺度因子曲线，图4最下方的曲线为5#故障电芯的样本熵-尺度因子曲线。根据图4得到图5所示的每条样本熵-尺度因子曲线的斜率。可以看出，5#故障电芯的样本熵-尺度因子曲线的斜率明显低于其它电芯，表明样本熵-尺度因子曲线的斜率可以体现电芯故障。
经过实验得到如图6所示的6个电芯的欧式距离，可以看出，5#故障电芯的欧式距离明显大于其它电芯，表面电芯充电电压曲线与平均电压曲线的欧式距离也可以体现电芯故障。
基于上文中的理论，本发明实施例提出了一种电池电芯故障诊断方法，如图7所示，包括：
步骤S1，获取多个采样时刻采集的电池多个电芯的充电电压；
步骤S2，根据多个充电电压计算待测电芯的MN绝对值之和以及欧氏距离；
步骤S3，根据多个采样时刻计算待测电芯的样本熵-尺度因子曲线斜率；
步骤S4，获取预设的正常电芯聚类中心和故障电芯聚类中心；
步骤S5，根据待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、正常电芯聚类中心计算第一汉明贴近度；
步骤S6，根据待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、故障电芯聚类中心计算第二汉明贴近度；
步骤S7，若第二汉明贴近度大于第一汉明贴近度，则判断待测电芯故障。
步骤S1中，假设每1s采集一次电池多个电芯的充电电压，共采集20s，则多个采样时刻分别为1、2、3、...、19、20。
步骤S2中，根据多个充电电压计算MN绝对值之和，包括：确定采样时刻采集的多个充电电压中的最大值；计算该采样时刻采集的多个充电电压的平均值；根据该采样时刻采集的待测电芯的充电电压、最大值和平均值计算该采样时刻对应的MN值；计算多个采样时刻对应的多个MN值的绝对值之和，得到MN绝对值之和。其中，根据该采样时刻采集的待测电芯的充电电压、最大值和平均值计算该采样时刻对应的MN值，包括：MN= ；MN为某个采样时刻对应的MN值，U为该采样时刻采集的待测电芯的充电电压， 为该采样时刻采集的所有电芯充电电压的平均值， 为该采样时刻采集的所有电芯充电电压中的最大值。这样可以得到单个采集时刻对应的待测电芯的MN绝对值之和，每个采集时刻均对应一个MN绝对值之和，可以得到所有采集时刻对应的所有MN绝对值之和。对MN值取绝对值，可以防止某些特殊情况下电芯的MN值在正负值之间来回跳动，从而导致取和后很小突显不出故障。
步骤S2中，根据多个充电电压计算欧氏距离，包括：计算采样时刻采集的多个充电电压的平均值；计算该采样时刻采集的待测电芯的充电电压减去平均值之后的差值；根据多个采样时刻对应的多个差值计算欧氏距离。其中，根据多个采样时刻对应的多个差值计算欧氏距离，包括：d1= ；d1为欧氏距离， 为采样时刻的序号，n为采样时刻的个数（例如20）， 为第 个采样时刻采集的待测电芯的充电电压， 为第 个采样时刻采集的所有电芯充电电压的平均值。
步骤S3包括：根据多个不同的尺度因子对多个采样时刻分别进行下采样，得到每个尺度因子下的多个采样时间；根据尺度因子下的多个采样时间计算该尺度因子对应的样本熵；对多个尺度因子以及对应的多个样本熵进行拟合，得到样本熵-尺度因子曲线；确定样本熵-尺度因子曲线的斜率。
其中，根据多个不同的尺度因子对多个采样时刻分别进行下采样，得到每个尺度因子下的多个采样时间，包括： = ，1≤b≤n/ ；b为采样时间的序号， 为第b个采样时间， 为尺度因子， 为采样时刻的序号， 为第 个采样时刻，n为采样时刻的个数。例如n=20， =1时，1≤b≤20， = 、 = 、...、 = ，多个采样时间为原始的采样时刻； =2时，1≤b≤10， =（ + ）/2、 =（ + ）/2、...、 =（ + ）/2； =3时，1≤b≤20/3， =（ + + ）/3、 =（ + + ）/3、...、 =（ + ）/3。
其中，计算尺度因子对应的样本熵的过程如下：
设采样时间的个数为e，1≤b≤e，将e个采样时间 重构为e-m+1个m维的向量X(1)、X(2)、…、X(e-m+1)，X(b)={ ， ，…， }；例如m=2，则将e个采样时间 重构为e-1个2维的向量X(1)、X(2)、…、X(e-1)， =2时e=10，则X(1)={ ， }、X(2)={ ， }、...、X(9)={ ， }；
计算d2= [ ]；i,j=1,2,...e-m+1，i≠j；d2为两个向量对应元素的差值绝对值的最大值；
计算 = ，r为给定阈值，例如r=0.2std，std为采样时间的标准差；
计算B= ；
令f= ，计算 = 、A= ；
样本熵S= ， 取有限值时S= 。
对每个尺度因子 计算一次样本熵S，这样可以得到多个尺度因子 对应的多个样本熵S。
步骤S4中，正常电芯聚类中心包括正常电芯MN聚类中心、正常电芯欧氏距离聚类中心和正常电芯斜率聚类中心；故障电芯聚类中心包括故障电芯MN聚类中心、故障电芯欧氏距离聚类中心和故障电芯斜率聚类中心。步骤S4具体包括：获取多个训练样本电芯中每个训练样本电芯的MN绝对值之和、欧氏距离以及样本熵-尺度因子曲线斜率；对多个MN绝对值之和进行归一化，对多个欧氏距离进行归一化，对多个样本熵-尺度因子曲线斜率进行归一化；对归一化后的样本熵-尺度因子曲线斜率取绝对值，对归一化后的MN绝对值之和、欧氏距离和样本熵-尺度因子曲线斜率的绝对值赋权重；对赋权重后的多个MN绝对值之和进行k-means聚类，得到正常电芯MN聚类中心和故障电芯MN聚类中心；对赋权重后的多个欧氏距离进行k-means聚类，得到正常电芯欧氏距离聚类中心和故障电芯欧氏距离聚类中心；对赋权重后的多个样本熵-尺度因子曲线斜率的绝对值进行k-means聚类，得到正常电芯斜率聚类中心和故障电芯斜率聚类中心。
其中，分别用每个训练样本电芯代替待测电芯进行步骤S1-S3，可以得到每个训练样本电芯的MN绝对值之和、欧氏距离以及样本熵-尺度因子曲线斜率；可以通过Z-score归一化方法进行归一化；对归一化后的样本熵-尺度因子曲线斜率取绝对值是为了便于横向比较；样本熵-尺度因子曲线斜率的重要性高，可以对归一化后的MN绝对值之和、欧氏距离和样本熵-尺度因子曲线斜率的绝对值分别赋权重1:1:2。
以MN绝对值之和为例，k-means聚类的过程为：第一步，随机选择2个MN绝对值之和作为初始质心；第二步，将每个MN绝对值之和指派到距离最近的初始质心，形成2个簇；第三步，计算第一个初始质心所在簇中所有MN绝对值之和的平均值并作为该簇的新的质心，计算第二个初始质心所在簇中所有MN绝对值之和的平均值并作为该簇的新的质心；第四步，重复第二步和第三步，直至2个簇的质心不再发生变化，2个簇的质心即为正常电芯MN聚类中心和故障电芯MN聚类中心。欧氏距离、样本熵-尺度因子曲线斜率的k-means聚类过程与MN绝对值之和相同。这里k-means聚类的目标函数为SSE。
步骤S5包括：N1=1- ；N1为第一汉明贴近度， 、 、 依次为待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率， 、 、 依次为正常电芯MN聚类中心、正常电芯欧氏距离聚类中心、正常电芯斜率聚类中心。
步骤S6包括：N2=1- ；N2为第二汉明贴近度， 、 、 依次为待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率， 、 、 依次为故障电芯MN聚类中心、故障电芯欧氏距离聚类中心、故障电芯斜率聚类中心。
本发明实施例中，若待测电芯的第二汉明贴近度不大于第一汉明贴近度，则判断待测电芯正常。
由上文可知，本发明实施例的电池电芯故障诊断方法根据电池电芯的充电电压计算待测电芯的MN绝对值之和以及欧氏距离，根据充电电压采样时刻计算待测电芯的样本熵-尺度因子曲线斜率，计算待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率与正常电芯聚类中心的第一汉明贴近度，计算待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率与故障电芯聚类中心的第二汉明贴近度，若第二汉明贴近度大于第一汉明贴近度则判断待测电芯故障，从三个维度综合诊断了电池电芯故障。
如图8所示，本发明实施例还提供一种电池电芯故障诊断装置，包括：
获取模块，用于获取多个采样时刻采集的电池多个电芯的充电电压；
计算模块，用于根据多个充电电压计算待测电芯的MN绝对值之和以及欧氏距离；
根据多个采样时刻计算待测电芯的样本熵-尺度因子曲线斜率；
获取模块，还用于获取预设的正常电芯聚类中心和故障电芯聚类中心；
计算模块，还用于根据待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、正常电芯聚类中心计算第一汉明贴近度；
根据待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、故障电芯聚类中心计算第二汉明贴近度；
判断模块，用于若第二汉明贴近度大于第一汉明贴近度，则判断待测电芯故障。
进一步的，计算模块，还可以用于：确定采样时刻采集的多个充电电压中的最大值；计算该采样时刻采集的多个充电电压的平均值；根据该采样时刻采集的待测电芯的充电电压、最大值和平均值计算该采样时刻对应的MN值；计算多个采样时刻对应的多个MN值的绝对值之和，得到MN绝对值之和。
进一步的，计算模块，还可以用于：计算采样时刻采集的多个充电电压的平均值；计算该采样时刻采集的待测电芯的充电电压减去平均值之后的差值；根据多个采样时刻对应的多个差值计算欧氏距离。
进一步的，计算模块，还可以用于：根据多个不同的尺度因子对多个采样时刻分别进行下采样，得到每个尺度因子下的多个采样时间；根据尺度因子下的多个采样时间计算该尺度因子对应的样本熵；对多个尺度因子以及对应的多个样本熵进行拟合，得到样本熵-尺度因子曲线；确定样本熵-尺度因子曲线的斜率。
进一步的，正常电芯聚类中心包括正常电芯MN聚类中心、正常电芯欧氏距离聚类中心和正常电芯斜率聚类中心；故障电芯聚类中心包括故障电芯MN聚类中心、故障电芯欧氏距离聚类中心和故障电芯斜率聚类中心；
获取模块，还可以用于：获取多个训练样本电芯中每个训练样本电芯的MN绝对值之和、欧氏距离以及样本熵-尺度因子曲线斜率；对多个MN绝对值之和进行归一化，对多个欧氏距离进行归一化，对多个样本熵-尺度因子曲线斜率进行归一化；对归一化后的样本熵-尺度因子曲线斜率取绝对值，对归一化后的MN绝对值之和、欧氏距离和样本熵-尺度因子曲线斜率的绝对值赋权重；对赋权重后的多个MN绝对值之和进行k-means聚类，得到正常电芯MN聚类中心和故障电芯MN聚类中心；对赋权重后的多个欧氏距离进行k-means聚类，得到正常电芯欧氏距离聚类中心和故障电芯欧氏距离聚类中心；对赋权重后的多个样本熵-尺度因子曲线斜率的绝对值进行k-means聚类，得到正常电芯斜率聚类中心和故障电芯斜率聚类中心。
进一步的，计算模块根据待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、正常电芯聚类中心计算第一汉明贴近度，可以包括：
N1=1- ；
N1为第一汉明贴近度， 、 、 依次为待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率， 、 、 依次为正常电芯MN聚类中心、正常电芯欧氏距离聚类中心、正常电芯斜率聚类中心。
进一步的，计算模块根据待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、故障电芯聚类中心计算第二汉明贴近度，可以包括：
N2=1- ；
N2为第二汉明贴近度， 、 、 依次为待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率， 、 、 依次为故障电芯MN聚类中心、故障电芯欧氏距离聚类中心、故障电芯斜率聚类中心。
基于与前文的电池电芯故障诊断方法同样的发明构思，本发明实施例还提供一种计算机设备，包括存储器、处理器及存储在存储器上的计算机程序，该处理器执行该计算机程序以实现前文任一项电池电芯故障诊断方法的步骤。
其中，总线架构（用总线来代表），总线可以包括任意数量的互联的总线和桥，总线将包括由处理器代表的一个或多个处理器和存储器代表的存储器的各种电路链接在一起。总线还可以将诸如外围设备、稳压器和功率管理电路等之类的各种其他电路链接在一起，这些都是本领域所公知的，因此，本文不再对其进行进一步描述。总线接口在总线和接收器和发送器之间提供接口。接收器和发送器可以是同一个元件，即收发机，提供用于在传输介质上与各种其他装置通信的单元。处理器负责管理总线和通常的处理，而存储器可以被用于存储处理器在执行操作时所使用的数据。
由于本发明实施例所介绍的计算机设备为实施本发明实施例中电池电芯故障诊断方法所采用的计算机设备，故而基于本发明实施例中所介绍的电池电芯故障诊断方法，本领域所属技术人员能够了解本发明实施例的计算机设备的具体实施方式以及其各种变化形式，所以在此对于该计算机设备如何实现本发明实施例中的方法不再详细介绍。只要本领域所属技术人员实施本发明实施例中电池电芯故障诊断方法所采用的计算机设备，都属于本发明所欲保护的范围。
基于与上述电池电芯故障诊断方法同样的发明构思，本发明还提供一种计算机可读存储介质，其上存储有计算机程序，该计算机程序被处理器执行时实现前文任一电池电芯故障诊断方法的步骤。
本领域内的技术人员应明白，本发明的实施例可提供为方法、系统、或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质（包括但不限于磁盘存储器、CD-ROM、光学存储器等）上实施的计算机程序产品的形式。
本发明是参照根据本发明实施例的方法、设备（系统）、和计算机程序产品的流程图和／或方框图来描述的。应理解可由计算机程序指令实现流程图和／或方框图中的每一流程和／或方框、以及流程图和／或方框图中的流程和／或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和／或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和／或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和／或方框图一个方框或多个方框中指定的功能的步骤。
尽管已描述了本发明的优选实施例，但本领域内的技术人员一旦得知了基本创造性概念，则可对这些实施例作出另外的变更和修改。所以，所附权利要求意欲解释为包括优选实施例以及落入本发明范围的所有变更和修改。
显然，本领域的技术人员可以对本发明进行各种改动和变型而不脱离本发明的精神和范围。这样，倘若本发明的这些修改和变型属于本发明权利要求及其等同技术的范围之内，则本发明也意图包含这些改动和变型在内。

Claims (10)

  1. 一种电池电芯故障诊断方法，应用于电池的诊断设备中，包括：
    获取多个采样时刻采集的电池多个电芯的充电电压；
    根据多个所述充电电压计算待测电芯的MN绝对值之和以及欧氏距离；
    根据多个所述采样时刻计算所述待测电芯的样本熵-尺度因子曲线斜率；
    获取预设的正常电芯聚类中心和故障电芯聚类中心；
    根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述正常电芯聚类中心计算第一汉明贴近度；
    根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述故障电芯聚类中心计算第二汉明贴近度；
    若所述第二汉明贴近度大于所述第一汉明贴近度，则判断所述待测电芯故障。
  2. 根据权利要求1所述的电池电芯故障诊断方法，其中，根据多个所述充电电压计算所述MN绝对值之和，包括：
    确定所述采样时刻采集的多个所述充电电压中的最大值；
    计算该采样时刻采集的多个所述充电电压的平均值；
    根据该采样时刻采集的所述待测电芯的所述充电电压、所述最大值和所述平均值计算该采样时刻对应的MN值；
    计算多个所述采样时刻对应的多个所述MN值的绝对值之和，得到所述MN绝对值之和。
  3. 根据权利要求1所述的电池电芯故障诊断方法，其中，根据多个所述充电电压计算所述欧氏距离，包括：
    计算所述采样时刻采集的多个所述充电电压的平均值；
    计算该采样时刻采集的所述待测电芯的所述充电电压减去所述平均值之后的差值；
    根据多个所述采样时刻对应的多个所述差值计算所述欧氏距离。
  4. 根据权利要求1所述的电池电芯故障诊断方法，其中，根据多个所述采样时刻计算所述待测电芯的样本熵-尺度因子曲线斜率，包括：
    根据多个不同的尺度因子对多个所述采样时刻分别进行下采样，得到每个所述尺度因子下的多个采样时间；
    根据所述尺度因子下的多个所述采样时间计算该尺度因子对应的样本熵；
    对多个所述尺度因子以及对应的多个所述样本熵进行拟合，得到样本熵-尺度因子曲线；
    确定所述样本熵-尺度因子曲线的斜率。
  5. 根据权利要求1所述的电池电芯故障诊断方法，其中，所述正常电芯聚类中心包括正常电芯MN聚类中心、正常电芯欧氏距离聚类中心和正常电芯斜率聚类中心；所述故障电芯聚类中心包括故障电芯MN聚类中心、故障电芯欧氏距离聚类中心和故障电芯斜率聚类中心；
    获取预设的正常电芯聚类中心和故障电芯聚类中心，包括：
    获取多个训练样本电芯中每个所述训练样本电芯的所述MN绝对值之和、所述欧氏距离以及所述样本熵-尺度因子曲线斜率；
    对多个所述MN绝对值之和进行归一化，对多个所述欧氏距离进行归一化，对多个所述样本熵-尺度因子曲线斜率进行归一化；
    对归一化后的所述样本熵-尺度因子曲线斜率取绝对值，对归一化后的所述MN绝对值之和、所述欧氏距离和所述样本熵-尺度因子曲线斜率的绝对值赋权重；
    对赋权重后的多个所述MN绝对值之和进行k-means聚类，得到所述正常电芯MN聚类中心和所述故障电芯MN聚类中心；
    对赋权重后的多个所述欧氏距离进行k-means聚类，得到所述正常电芯欧氏距离聚类中心和所述故障电芯欧氏距离聚类中心；
    对赋权重后的多个所述样本熵-尺度因子曲线斜率的绝对值进行k-means聚类，得到所述正常电芯斜率聚类中心和所述故障电芯斜率聚类中心。
  6. 根据权利要求5所述的电池电芯故障诊断方法，其中，根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述正常电芯聚类中心计算第一汉明贴近度，包括：
    N1=1- ；
    N1为所述第一汉明贴近度， 、 、 依次为所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率， 、 、 依次为所述正常电芯MN聚类中心、所述正常电芯欧氏距离聚类中心、所述正常电芯斜率聚类中心。
  7. 根据权利要求5所述的电池电芯故障诊断方法，其中，根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述故障电芯聚类中心计算第二汉明贴近度，包括：
    N2=1- ；
    N2为所述第二汉明贴近度， 、 、 依次为所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率， 、 、 依次为所述故障电芯MN聚类中心、所述故障电芯欧氏距离聚类中心、所述故障电芯斜率聚类中心。
  8. 一种电池电芯故障诊断装置，包括：
    获取模块，用于获取多个采样时刻采集的电池多个电芯的充电电压；
    计算模块，用于根据多个所述充电电压计算待测电芯的MN绝对值之和以及欧氏距离；
    根据多个所述采样时刻计算所述待测电芯的样本熵-尺度因子曲线斜率；
    所述获取模块，还用于获取预设的正常电芯聚类中心和故障电芯聚类中心；
    所述计算模块，还用于根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述正常电芯聚类中心计算第一汉明贴近度；
    根据所述待测电芯的MN绝对值之和、欧氏距离、样本熵-尺度因子曲线斜率、所述故障电芯聚类中心计算第二汉明贴近度；
    判断模块，用于若所述第二汉明贴近度大于所述第一汉明贴近度，则判断所述待测电芯故障。
  9. 一种计算机设备，包括存储器、处理器及存储在存储器上的计算机程序，所述处理器执行所述计算机程序以实现权利要求1-7中任一项权利要求所述方法的步骤。
  10. 一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现权利要求1-7中任一项权利要求所述方法的步骤。
PCT/CN2024/138428 2024-06-21 2024-12-11 电池电芯故障诊断方法及装置 Pending WO2025260641A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202410809826.6 2024-06-21
CN202410809826.6A CN118688672B (zh) 2024-06-21 2024-06-21 电池电芯故障诊断方法及装置

Publications (1)

Publication Number Publication Date
WO2025260641A1 true WO2025260641A1 (zh) 2025-12-26

Family

ID=92776069

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2024/138428 Pending WO2025260641A1 (zh) 2024-06-21 2024-12-11 电池电芯故障诊断方法及装置

Country Status (2)

Country Link
CN (1) CN118688672B (zh)
WO (1) WO2025260641A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118688672B (zh) * 2024-06-21 2026-04-14 东风汽车集团股份有限公司 电池电芯故障诊断方法及装置

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190033386A1 (en) * 2017-07-31 2019-01-31 Lg Chem, Ltd. Self-diagnosing battery cell monitoring system
CN110045298A (zh) * 2019-05-06 2019-07-23 重庆大学 一种动力电池组参数不一致性的诊断方法
CN110703109A (zh) * 2019-10-25 2020-01-17 山东大学 一种基于修正样本熵的电池串多故障诊断方法及系统
WO2022151819A1 (zh) * 2021-01-18 2022-07-21 北京理工大学 一种基于聚类分析的电池系统在线故障诊断方法和系统
CN115144772A (zh) * 2022-07-26 2022-10-04 合肥工业大学智能制造技术研究院 一种锂离子电池模组的虚焊检测方法
WO2023274318A1 (zh) * 2021-06-29 2023-01-05 蜂巢能源科技股份有限公司 电池突发型内短路诊断方法、装置、存储介质及电子设备
CN117572268A (zh) * 2023-11-07 2024-02-20 合肥力高动力科技有限公司 一种基于ic曲线检测动力电池内短路的方法
CN118688672A (zh) * 2024-06-21 2024-09-24 东风汽车集团股份有限公司 电池电芯故障诊断方法及装置

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112014759B (zh) * 2020-07-22 2023-04-14 量道(深圳)储能科技有限公司 基于分布式集群分析电池性能的方法及系统
KR102904761B1 (ko) * 2022-10-31 2025-12-29 주식회사 에스앤에스 배터리 셀 모니터링 장치 및 방법
CN115774208A (zh) * 2022-11-25 2023-03-10 上海玫克生储能科技有限公司 一种电池单元异常监测方法、装置、电子设备及介质
CN116027212B (zh) * 2023-02-10 2025-09-02 上海电享信息科技有限公司 基于多方法融合的电芯异常识别方法、装置和电子设备
CN117517966A (zh) * 2023-10-23 2024-02-06 华东理工大学 一种电池组的故障诊断方法
CN117991118A (zh) * 2024-02-02 2024-05-07 阳光新能源开发股份有限公司 异常电芯检测方法、装置、设备及存储介质
CN118209870B (zh) * 2024-05-22 2024-08-02 深圳市泰科动力系统有限公司 一种锂电池健康水平的评估管理方法及系统

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190033386A1 (en) * 2017-07-31 2019-01-31 Lg Chem, Ltd. Self-diagnosing battery cell monitoring system
CN110045298A (zh) * 2019-05-06 2019-07-23 重庆大学 一种动力电池组参数不一致性的诊断方法
CN110703109A (zh) * 2019-10-25 2020-01-17 山东大学 一种基于修正样本熵的电池串多故障诊断方法及系统
WO2022151819A1 (zh) * 2021-01-18 2022-07-21 北京理工大学 一种基于聚类分析的电池系统在线故障诊断方法和系统
WO2023274318A1 (zh) * 2021-06-29 2023-01-05 蜂巢能源科技股份有限公司 电池突发型内短路诊断方法、装置、存储介质及电子设备
CN115144772A (zh) * 2022-07-26 2022-10-04 合肥工业大学智能制造技术研究院 一种锂离子电池模组的虚焊检测方法
CN117572268A (zh) * 2023-11-07 2024-02-20 合肥力高动力科技有限公司 一种基于ic曲线检测动力电池内短路的方法
CN118688672A (zh) * 2024-06-21 2024-09-24 东风汽车集团股份有限公司 电池电芯故障诊断方法及装置

Also Published As

Publication number Publication date
CN118688672A (zh) 2024-09-24
CN118688672B (zh) 2026-04-14

Similar Documents

Publication Publication Date Title
CN116936978B (zh) 一种蓄电池鼓包智能化监测方法及系统
CN105488539B (zh) 分类模型的生成方法及装置、系统容量的预估方法及装置
CN114662386B (zh) 一种轴承故障诊断方法及系统
WO2016107246A1 (zh) 基于小波降噪和相关向量机的锂电池剩余寿命预测方法
CN116664928B (zh) 一种基于CNN与Transformer的糖尿病视网膜病变分级方法与系统
US20190258935A1 (en) Computer-readable recording medium, learning method, and learning apparatus
CN109597392A (zh) 有助于故障诊断的方法、装置和设备以及机器可读介质
CN112200114A (zh) 故障诊断模型的训练方法、装置、电子设备及存储介质
WO2025260641A1 (zh) 电池电芯故障诊断方法及装置
CN117435969A (zh) 开关柜健康状态评估方法、装置、设备及存储介质
CN108254689A (zh) 电池组反极单体电池检测方法和系统
CN115877222A (zh) 一种储能电站故障检测方法、装置、介质及储能电站
CN117150401A (zh) 异常检测方法、装置、计算机设备、存储介质和程序产品
CN117799440A (zh) 动力电池的一致性预警方法、装置和计算机设备
CN120103200B (zh) 基于融合特征的锂电池soh预测方法、装置、设备和介质
US20120323518A1 (en) Methods systems and apparatus for ranking tests used to identify faults in a system
CN120522597A (zh) 一种面向智能运维的电池组健康评估方法
CN119716552A (zh) 用于检测电池包中异常单体电芯的方法和装置
CN117251714A (zh) 设备状态评估方法、处理器及系统
CN116719701A (zh) 储能系统运行状态的确定方法、装置及计算机设备
CN114740389A (zh) 电池健康评估方法、装置、电子设备和可读存储介质
Zhang et al. Fault diagnosis for battery pack using SDP-DRN method with long-term operational data
CN116028556A (zh) 一种ar终端系统传输数据智能处理方法
Zhang et al. Bearing performance degradation assessment based on a combination of multi-scale entropy and K-medoids clustering
CN114021925B (zh) 电力系统的安全评估方法、装置、计算机设备和存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 24945368

Country of ref document: EP

Kind code of ref document: A1