WO2022151819A1 - 一种基于聚类分析的电池***在线故障诊断方法和*** - Google Patents

一种基于聚类分析的电池***在线故障诊断方法和*** Download PDF

Info

Publication number
WO2022151819A1
WO2022151819A1 PCT/CN2021/129524 CN2021129524W WO2022151819A1 WO 2022151819 A1 WO2022151819 A1 WO 2022151819A1 CN 2021129524 W CN2021129524 W CN 2021129524W WO 2022151819 A1 WO2022151819 A1 WO 2022151819A1
Authority
WO
WIPO (PCT)
Prior art keywords
battery cell
battery
cluster
voltage
battery cells
Prior art date
Application number
PCT/CN2021/129524
Other languages
English (en)
French (fr)
Inventor
王震坡
孙振宇
刘鹏
张照生
逄昊
尹豪
Original Assignee
北京理工大学
北京理工新源信息科技有限公司
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 北京理工大学, 北京理工新源信息科技有限公司 filed Critical 北京理工大学
Publication of WO2022151819A1 publication Critical patent/WO2022151819A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/367Software therefor, e.g. for battery testing using modelling or look-up tables
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/36Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
    • G01R31/385Arrangements for measuring battery or accumulator variables
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02TCLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00Road transport of goods or passengers
    • Y02T10/60Other road transportation technologies with climate change mitigation effect
    • Y02T10/70Energy storage systems for electromobility, e.g. batteries

Definitions

  • the invention relates to the field of battery cell fault detection, in particular to an on-line fault diagnosis method and system for a battery system based on cluster analysis.
  • Lithium-ion batteries have become the mainstream energy storage devices for electric vehicles due to their long cycle life, high voltage, high output power, and low pollution.
  • the lithium-ion power battery system is usually composed of many battery cells connected in series and parallel.
  • each battery cell connected in series has the same current excitation. have the same trend of voltage change.
  • the temperature is also different.
  • the difference between the battery cells will be more obvious, resulting in obvious inconsistency in the change of the voltage of each battery cell.
  • the new energy vehicle battery management system can obtain data such as battery temperature and voltage.
  • the battery model and real vehicle data are used to generate residuals of battery characteristics such as voltage and temperature, and the residuals are used to determine whether the battery is faulty.
  • This method requires high accuracy of the battery model and is difficult to implement in real vehicles. application to realize online diagnosis.
  • the data-driven fault diagnosis method does not require an accurate battery model, it requires a large amount of sample data for training, and the computational cost is high.
  • the present invention provides an online fault diagnosis method and system for a battery system based on cluster analysis.
  • the present invention provides the following scheme:
  • An online fault diagnosis method for a battery system based on cluster analysis comprising:
  • the operation data includes: the voltage, current and temperature of each battery cell;
  • a voltage matrix is formed according to the operating data; the rows of the voltage matrix represent battery cell serial numbers, and the columns of the voltage matrix represent time series;
  • the battery cells in the electric vehicle are divided into abnormal battery cell clusters and normal battery cell clusters according to the voltage matrix;
  • the relevant parameters include: correlation coefficient and fluctuation variance
  • the preset threshold includes: a quantity ratio threshold and a Euclidean distance threshold;
  • an abnormal battery cell is determined, and the serial number of the abnormal battery cell is output.
  • the K-means clustering algorithm is used to divide the battery cells in the electric vehicle into abnormal battery cell clusters and normal battery cell clusters according to the voltage matrix, specifically including:
  • a sample set is constructed according to the voltage matrix; the sample set includes: a plurality of elements composed of the correlation coefficient of each battery cell and the fluctuation variance of each battery cell;
  • the K-means clustering algorithm is used to divide the battery cells in the electric vehicle into abnormal battery cell clusters and normal battery cell clusters based on the sample set.
  • the constructing the sample set according to the voltage matrix specifically includes:
  • the sample set is constructed according to the correlation coefficient of the battery cells and the fluctuation variance of the battery cells.
  • the determining the fluctuation variance of the battery cell according to the voltage value and the voltage mean value specifically includes:
  • the fluctuation variance of the battery cell is determined according to the trending vector.
  • the present invention discloses the following technical effects:
  • the on-line fault diagnosis method for a battery system based on cluster analysis uses K-means clustering algorithm to classify the battery cells in the battery system of the electric vehicle based on the obtained operation data of the electric vehicle, and then obtains the result according to the classification.
  • K-means clustering algorithm to classify the battery cells in the battery system of the electric vehicle based on the obtained operation data of the electric vehicle, and then obtains the result according to the classification.
  • the Euclidean distance between the two battery cell clusters can quickly and accurately determine the abnormal battery cells, and output the battery cell serial numbers to reduce the difficulty of battery cell fault monitoring in real vehicles.
  • the present invention also provides an on-line battery system fault diagnosis system based on the cluster analysis.
  • the cluster analysis-based online fault diagnosis system for battery systems includes:
  • an operation data acquisition module for acquiring the operation data of the electric vehicle;
  • the operation data includes: the voltage, current and temperature of each battery cell;
  • a voltage matrix forming module configured to form a voltage matrix according to the operating data; the rows of the voltage matrix represent battery cell serial numbers, and the columns of the voltage matrix represent time series;
  • the cluster classification module is used for using the K-means clustering algorithm to divide the battery cells in the electric vehicle into abnormal battery cell clusters and normal battery cell clusters according to the voltage matrix;
  • a parameter determination module configured to determine the number ratio of the battery cells in the abnormal battery cell cluster and the normal battery cell cluster, and respectively determine the cluster center in the abnormal battery cell cluster and the normal battery cell
  • the relevant parameters of the cluster center in the cluster include: correlation coefficient and fluctuation variance
  • an Euclidean distance determining module configured to determine the Euclidean distance between the cluster center of the abnormal battery cell cluster and the cluster center of the normal battery cell cluster according to the relevant parameter
  • a threshold acquisition module used for acquiring a preset threshold
  • the preset threshold includes: a quantity ratio threshold and a Euclidean distance threshold
  • An abnormal battery cell determination module configured to determine an abnormal battery cell according to the relationship between the number ratio and the number ratio threshold, and the relationship between the Euclidean distance and the Euclidean distance threshold, and output the abnormal battery The serial number of the unit.
  • the cluster classification module specifically includes:
  • a sample set construction sub-module configured to construct a sample set according to the voltage matrix;
  • the sample set includes: a plurality of elements formed by the correlation coefficient of each battery cell and the fluctuation variance of each battery cell;
  • the cluster classification submodule is configured to use the K-means clustering algorithm to classify the battery cells in the electric vehicle into abnormal battery cell clusters and normal battery cell clusters based on the sample set.
  • the sample set construction submodule specifically includes:
  • a Pearson correlation coefficient determination unit configured to determine the Pearson correlation coefficient between two adjacent battery cells in the voltage matrix
  • a correlation coefficient determination unit used for determining the correlation coefficient of the battery cells according to the determined Pearson correlation coefficient between two adjacent battery cells
  • a voltage value acquisition unit used to acquire the voltage value of each battery cell and the voltage average value of all battery cells
  • a fluctuation variance determining unit configured to determine the fluctuation variance of the battery cells according to the voltage value of each battery cell and the voltage average value of all the battery cells;
  • a sample set construction unit configured to construct the sample set according to the correlation coefficient of the battery cells and the fluctuation variance of the battery cells.
  • the fluctuation variance determination unit specifically includes:
  • a trending vector determining subunit configured to obtain a trending vector after trending the battery cells according to the voltage value of each battery cell and the voltage average value of all the battery cells;
  • the fluctuation variance determination subunit is used for determining the fluctuation variance of the battery cell according to the trending vector.
  • the technical effect achieved by the cluster analysis-based battery system online fault diagnosis system provided by the present invention is the same as the technical effect achieved by the cluster analysis-based battery system online fault diagnosis method provided by the present invention, and will not be repeated here.
  • Fig. 1 is the flow chart of the battery system online fault diagnosis method based on cluster analysis provided by the present invention
  • FIG. 2 is a schematic diagram of data collection provided by an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of a battery cell correlation coefficient calculation principle provided by an embodiment of the present invention.
  • FIG. 4 is a schematic structural diagram of an on-line fault diagnosis system for a battery system based on cluster analysis provided by the present invention.
  • the on-line fault diagnosis method and system for a battery system based on cluster analysis provided by the present invention perform clustering identification of faulty battery cells by calculating the correlation coefficient and fluctuation variance of the battery cells, and has the advantages of fast calculation speed, high accuracy and easy implementation in practice. It is used in the car for on-line fault diagnosis of the battery and other advantages.
  • Fig. 1 is the flow chart of the battery system online fault diagnosis method based on cluster analysis provided by the present invention, as shown in Fig. 1, a kind of battery system online fault diagnosis method based on cluster analysis, including:
  • Step 100 Obtain the operation data of the electric vehicle.
  • Operational data includes: voltage, current and temperature of each battery cell.
  • the process of obtaining operating data in the present invention is: collecting the operating data (such as voltage, current and temperature) of the electric vehicle obtained from multiple sensors into a transmitting terminal box (T-BOX). All data is transmitted to the storage server of the new energy vehicle big data platform through the wireless network (4G or 5G), which follows the transmission protocol named by the Electric Vehicle Remote Service and Management System (GB/T 32960).
  • Step 101 Form a voltage matrix according to the operating data.
  • the rows of the voltage matrix represent battery cell serial numbers, and the columns of the voltage matrix represent the time series.
  • the collected operation data needs to be preprocessed.
  • the preprocessing process is as follows: de-duplication of multiple frames of data collected at the same time point, and only one frame of data is retained. When a certain frame of battery cell voltage lacks more than two, the data frame is deleted.
  • the n battery cells at time t are expressed as a vector
  • the battery cell voltage vector from time t 0 to time t s can be expressed as Therefore, the voltage matrix of all battery cells can be expressed as
  • Step 102 Using the K-means clustering algorithm, according to the voltage matrix, the battery cells in the electric vehicle are divided into abnormal battery cell clusters and normal battery cell clusters. Specifically, this step includes:
  • Step 1021 Construct a sample set according to the voltage matrix.
  • the sample set includes: a plurality of elements consisting of the correlation coefficient of each battery cell and the fluctuation variance of each battery cell.
  • the construction process of this sample set is as follows:
  • Step 10211 Determine the Pearson correlation coefficient between two adjacent battery cells in the voltage matrix.
  • the specific principle of determining the Pearson correlation coefficient is shown in Figure 3, and the process is as follows:
  • the original formula for the coefficients is:
  • ⁇ x and ⁇ y are the standard deviations of the vectors x and y, respectively, and cov(x, y) is the covariance of the x and y vectors.
  • V 1 , V 2 , . . . V D in FIG. 3 all represent the voltage values of the battery cells.
  • Step 10212 Determine the correlation coefficient of the battery cells according to the determined Pearson correlation coefficient between two adjacent battery cells.
  • the specific calculation formula of the correlation coefficient is:
  • Step 10213 Obtain the voltage value of each battery cell and the voltage average value of all battery cells.
  • Step 10214 Determine the fluctuation variance of the battery cells according to the voltage value of each battery cell and the voltage average value of all the battery cells.
  • the determination process of the fluctuation variance is:
  • a trend vector is obtained. Specifically, for a certain frame, the voltage average value of all battery cells in the frame is firstly obtained, and then the voltage value of each battery cell in the frame is subtracted from the voltage average value of all battery cells in the frame, which is denoted as DV .
  • the trending vector formed after detrending is expressed as:
  • the fluctuation variance of the cell is determined from the trending vector. Specifically, the fluctuation variance of the battery cell i in the time period t 0 -t s is calculated according to the following formula:
  • Step 10215 Construct a sample set according to the correlation coefficient of the battery cells and the fluctuation variance of the battery cells.
  • Step 1022 using the K-means clustering algorithm to divide the battery cells in the electric vehicle into abnormal battery cell clusters and normal battery cell clusters based on the sample set.
  • the process of clustering is as follows:
  • the clusters of clustering are divided into two categories, namely h-2, and the center of each cluster is denoted by c 1 and c 2 , respectively.
  • Two battery cells qi and q j are randomly selected as cluster centers among all battery cells.
  • the cluster centers are recalculated using the following formula:
  • the numerator in formula (7) represents the sum of all values in each cluster, and the denominator represents the number of samples in each cluster.
  • Step 103 Determine the number ratio of the battery cells in the abnormal battery cell cluster and the normal battery cell cluster, and determine the relevant parameters of the cluster center in the abnormal battery cell cluster and the cluster center in the normal battery cell cluster respectively. Relevant parameters include: correlation coefficient and volatility variance.
  • the number ratio of battery cells in a normal battery cell cluster is determined by the following formula:
  • ka is the number ratio
  • n abnormal and n normal represent the number of battery cells in the abnormal cell cluster and the normal cell cluster, respectively.
  • Step 104 Determine the Euclidean distance between the cluster center of the abnormal battery cell cluster and the cluster center of the normal battery cell cluster according to the relevant parameters.
  • the formula for calculating Euclidean distance is as follows:
  • Step 105 Obtain a preset threshold.
  • the preset thresholds include: the number ratio threshold ks and the Euclidean distance threshold ds .
  • Step 106 Determine the abnormal battery cell according to the relationship between the number ratio and the number ratio threshold, and the relationship between the Euclidean distance and the Euclidean distance threshold, and output the serial number of the abnormal battery cell. Specifically, if d> d s and ka ⁇ ks , h is set to 2, and the battery cell number in the abnormal cluster is alerted. Otherwise, h is set to 1 and no abnormal cell alarm is given.
  • the present invention also provides an on-line battery system fault diagnosis system based on the cluster analysis.
  • the battery system online fault diagnosis system based on cluster analysis includes: operation data acquisition module 1, voltage matrix formation module 2, cluster classification module 3, parameter determination module 4, Euclidean distance determination module 5, threshold value
  • the acquisition module 6 and the abnormal battery cell determination module 7 are obtained.
  • the operation data acquisition module 1 is used to acquire the operation data of the electric vehicle.
  • Operational data includes: voltage, current and temperature of each battery cell.
  • the voltage matrix forming module 2 is used for forming a voltage matrix according to the operation data.
  • the rows of the voltage matrix represent battery cell serial numbers, and the columns of the voltage matrix represent the time series.
  • the cluster classification module 3 is used for using the K-means clustering algorithm to divide the battery cells in the electric vehicle into abnormal battery cell clusters and normal battery cell clusters according to the voltage matrix.
  • the parameter determination module 4 is used to determine the number ratio of the battery cells in the abnormal battery cell cluster and the normal battery cell cluster, and respectively determine the relevant parameters of the cluster center in the abnormal battery cell cluster and the cluster center in the normal battery cell cluster. Relevant parameters include: correlation coefficient and volatility variance.
  • the Euclidean distance determining module 5 is used for determining the Euclidean distance between the cluster center of the abnormal battery cell cluster and the cluster center of the normal battery cell cluster according to the relevant parameters.
  • the threshold value obtaining module 6 is used for obtaining a preset threshold value.
  • the preset thresholds include: number ratio threshold and Euclidean distance threshold.
  • the abnormal battery cell determination module 7 is used to determine the abnormal battery cell according to the relationship between the number ratio and the number ratio threshold, and the relationship between the Euclidean distance and the Euclidean distance threshold, and output the serial number of the abnormal battery cell.
  • the above cluster classification module 3 specifically includes: a sample set construction submodule and a cluster classification submodule.
  • the sample set construction sub-module is used to construct the sample set according to the voltage matrix.
  • the sample set includes: a plurality of elements consisting of the correlation coefficient of each battery cell and the fluctuation variance of each battery cell.
  • the cluster classification submodule is used to classify the battery cells in the electric vehicle into abnormal battery cell clusters and normal battery cell clusters based on the sample set using the K-means clustering algorithm.
  • the sample set construction sub-module may include: a Pearson correlation coefficient determination unit, a correlation coefficient determination unit, a voltage value acquisition unit, a fluctuation variance determination unit, and a sample set construction unit.
  • the Pearson correlation coefficient determination unit is used to determine the Pearson correlation coefficient between two adjacent battery cells in the voltage matrix.
  • the correlation coefficient determination unit is configured to determine the correlation coefficient of the battery cells according to the determined Pearson correlation coefficient between two adjacent battery cells.
  • the voltage value acquiring unit is used for acquiring the voltage value of each battery cell and the voltage average value of all the battery cells.
  • the fluctuation variance determining unit is used for determining the fluctuation variance of the battery cells according to the voltage value of each battery cell and the voltage average value of all the battery cells.
  • the sample set construction unit is used to construct a sample set according to the correlation coefficient of the battery cells and the fluctuation variance of the battery cells.
  • the above-mentioned fluctuation variance determining unit specifically includes: a trending vector determining subunit and a fluctuation variance determining subunit.
  • the trending vector determination subunit is used to obtain a trending vector after trending the battery cells according to the voltage value of each battery cell and the voltage average value of all the battery cells.
  • the fluctuation variance determination subunit is used to determine the fluctuation variance of the battery cell according to the trending vector.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Secondary Cells (AREA)
  • Tests Of Electric Status Of Batteries (AREA)

Abstract

本发明涉及一种基于聚类分析的电池***在线故障诊断方法和***。本发明提供的基于聚类分析的电池***在线故障诊断方法和***,基于获取的电动汽车的运行数据,采用K-means聚类算法对电动汽车电池***中的电池单体进行簇分类,然后依据分类得到的两种电池单体簇间的欧式距离,快速、准确的确定异常的电池单体,并进行电池单体序号的输出,以降低实车中电池单体故障监测的难度。

Description

一种基于聚类分析的电池***在线故障诊断方法和*** 技术领域
本发明涉及电池单体故障检测领域,特别是涉及一种基于聚类分析的电池***在线故障诊断方法和***。
背景技术
锂离子电池因其循环寿命长、电压高、输出功率大、低污染的特性成为目前主流的电动汽车储能设备。为了获得足够的输出功率,锂离子动力电池***通常由许多电池单体电池通过串并联的方式组成,在车辆行驶中,串联起来的各个电池单体有着相同的电流激励,因此,在理论上应该有相同的电压变化趋势。但是由于组成电池包的电池单体在初始时在性能上就有一定差异,并且在实车中各个电池单体分布的位置,温度也不同,当电池长时间运行或者当电池受到碰撞挤压等外力影响时,电池单体之间的差异会更加明显,从而导致各个电池单体电压的变化产生明显的不一致性。
新能源汽车电池管理***能够获得电池温度与电压等数据。对于基于电池模型诊断方法利用电池模型与实车数据生成电压、温度等电池特征的残差,根据残差来判断电池是否发生故障,这种方法对于电池模型精度要求很高,难以在实车上应用,实现在线诊断。基于数据驱动的故障诊断方法虽然不需要精确的电池模型,但是需要大量的样本数据进行训练,计算成本高。
发明内容
为解决现有技术中存在的上述问题,本发明提供了一种基于聚类分析的电池***在线故障诊断方法和***。
为实现上述目的,本发明提供了如下方案:
一种基于聚类分析的电池***在线故障诊断方法,包括:
获取电动汽车的运行数据;所述运行数据包括:每一电池单体的电压、电流和温度;
根据所述运行数据形成电压矩阵;所述电压矩阵的行代表电池单体序号, 所述电压矩阵的列代表时间序列;
采用K-means聚类算法,根据所述电压矩阵将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇;
确定所述异常电池单体簇和所述正常电池单体簇中电池单体的数量比,并分别确定所述异常电池单体簇中簇中心和所述正常电池单体簇中簇中心的相关参数;所述相关参数包括:相关系数和波动方差;
根据所述相关参数确定所述异常电池单体簇的簇中心与所述正常电池单体簇的簇中心间的欧式距离;
获取预设阈值;所述预设阈值包括:数量比阈值和欧式距离阈值;
根据所述数量比与所述数量比阈值间的关系,以及所述欧式距离与所述欧式距离阈值间的关系确定异常电池单体,并输出所述异常电池单体的序号。
优选地,所述采用K-means聚类算法,根据所述电压矩阵将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇,具体包括:
根据所述电压矩阵构建样本集;所述样本集包括:多个由每一电池单体的相关系数和每一电池单体波动方差构成的元素;
采用所述K-means聚类算法基于所述样本集将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇。
优选地,所述根据所述电压矩阵构建样本集,具体包括:
确定所述电压矩阵中相邻两个电池单体间的皮尔森相关系数;
根据确定的相邻两个电池单体间的皮尔森相关系数确定电池单体的相关系数;
获取每一电池单体的电压值以及所有电池单体的电压均值;
根据所述每一电池单体的电压值以及所有电池单体的电压均值确定电池单体的波动方差;
根据所述电池单体的相关系数和所述电池单体的波动方差构建所述样本集。
优选地,所述根据所述电压值和电压均值确定电池单体的波动方差,具体包括:
根据所述每一电池单体的电压值以及所有电池单体的电压均值对所述电池单体进行趋势化处理后,得到趋势化向量;
根据所述趋势化向量确定电池单体的波动方差。
根据本发明提供的具体实施例,本发明公开了以下技术效果:
本发明提供的基于聚类分析的电池***在线故障诊断方法,基于获取的电动汽车的运行数据,采用K-means聚类算法对电动汽车电池***中的电池单体进行簇分类,然后依据分类得到的两种电池单体簇间的欧式距离,快速、准确的确定异常的电池单体,并进行电池单体序号的输出,以降低实车中电池单体故障监测的难度。
此外,对应于上述提供的基于聚类分析的电池***在线故障诊断方法,本发明还提供了一种基于聚类分析的电池***在线故障诊断***。该基于聚类分析的电池***在线故障诊断***,包括:
运行数据获取模块,用于获取电动汽车的运行数据;所述运行数据包括:每一电池单体的电压、电流和温度;
电压矩阵形成模块,用于根据所述运行数据形成电压矩阵;所述电压矩阵的行代表电池单体序号,所述电压矩阵的列代表时间序列;
簇分类模块,用于采用K-means聚类算法,根据所述电压矩阵将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇;
参数确定模块,用于确定所述异常电池单体簇和所述正常电池单体簇中电池单体的数量比,并分别确定所述异常电池单体簇中簇中心和所述正常电池单体簇中簇中心的相关参数;所述相关参数包括:相关系数和波动方差;
欧式距离确定模块,用于根据所述相关参数确定所述异常电池单体簇的簇中心与所述正常电池单体簇的簇中心间的欧式距离;
阈值获取模块,用于获取预设阈值;所述预设阈值包括:数量比阈值和欧式距离阈值;
异常电池单体确定模块,用于根据所述数量比与所述数量比阈值间的关系,以及所述欧式距离与所述欧式距离阈值间的关系确定异常电池单体,并输出所述异常电池单体的序号。
优选地,所述簇分类模块具体包括:
样本集构建子模块,用于根据所述电压矩阵构建样本集;所述样本集包括:多个由每一电池单体的相关系数和每一电池单体波动方差构成的元素;
簇分类子模块,用于采用所述K-means聚类算法基于所述样本集将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇。
优选地,所述样本集构建子模块具体包括:
皮尔森相关系数确定单元,用于确定所述电压矩阵中相邻两个电池单体间的皮尔森相关系数;
相关系数确定单元,用于根据确定的相邻两个电池单体间的皮尔森相关系数确定电池单体的相关系数;
电压值获取单元,用于获取每一电池单体的电压值以及所有电池单体的电压均值;
波动方差确定单元,用于根据所述每一电池单体的电压值以及所有电池单体的电压均值确定电池单体的波动方差;
样本集构建单元,用于根据所述电池单体的相关系数和所述电池单体的波动方差构建所述样本集。
优选地,所述波动方差确定单元具体包括:
趋势化向量确定子单元,用于根据所述每一电池单体的电压值以及所有电池单体的电压均值对所述电池单体进行趋势化处理后,得到趋势化向量;
波动方差确定子单元,用于根据所述趋势化向量确定电池单体的波动方差。
本发明提供的基于聚类分析的电池***在线故障诊断***实现的技术效果与上述本发明提供的基于聚类分析的电池***在线故障诊断方法实现的技术效果相同,在此不再进行赘述。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅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性的前提下,还可以根据这些附图获得其他的附图。
图1为本发明提供的基于聚类分析的电池***在线故障诊断方法的流程图;
图2为本发明实施例提供的数据采集的原理图;
图3为本发明实施例提供的电池单体相关系数计算原理图;
图4为本发明提供的基于聚类分析的电池***在线故障诊断***的结构示意图。
具体实施方式
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
本发明提供的基于聚类分析的电池***在线故障诊断方法和***,通过计算电池单体相关系数和波动方差进行聚类识别故障电池单体,具有计算速度快、准确率高以及易于实施在实车上用于电池在线故障诊断等优点。
为使本发明的上述目的、特征和优点能够更加明显易懂,下面结合附图和具体实施方式对本发明作进一步详细的说明。
图1为本发明提供的基于聚类分析的电池***在线故障诊断方法的流程图,如图1所示,一种基于聚类分析的电池***在线故障诊断方法,包括:
步骤100:获取电动汽车的运行数据。运行数据包括:每一电池单体的电压、电流和温度。具体的,如图2所示,本发明获取运行数据的过程为:将从多个传感器获得的电动汽车的运行数据(例如电压,电流和温度)收集到发射端子盒(T-BOX)中。所有数据通过无线网络(4G或5G)传输到新能源汽车大数据平台的存储服务器,该网络遵循电动汽车远程服务和管理***(GB/T  32960)命名的传输协议。
步骤101:根据运行数据形成电压矩阵。电压矩阵的行代表电池单体序号,电压矩阵的列代表时间序列。
具体的,在形成电压矩阵之前,需要对采集得到的运行数据进行预处理。预处理的过程为:对在相同时间点采集的多帧数据去重,只保留一帧数据。当某帧电池单体电压缺少两个以上时,删除该数据帧。
基于上述预处理后得到的数据构建电压矩阵的过程具体为:
首先将第t时刻n个电池单体表示为矢量
Figure PCTCN2021129524-appb-000001
从t 0时刻到t s时刻的电池单体电压向量可表示为
Figure PCTCN2021129524-appb-000002
因此,所有电池单体的电压矩阵可表示为
Figure PCTCN2021129524-appb-000003
步骤102:采用K-means聚类算法,根据电压矩阵将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇。具体的,该步骤包括:
步骤1021:根据电压矩阵构建样本集。样本集包括:多个由每一电池单体的相关系数和每一电池单体波动方差构成的元素。该样本集的构建过程具体为:
步骤10211:确定电压矩阵中相邻两个电池单体间的皮尔森相关系数。确定皮尔森相关系数的具体原理如图3所示,其过程具体为:
概念引入:
x、y为两个向量,x=[x 1,x 2,x 3,....,x N],y=[y 1,y 2,y 3,....,y N]相关系数的原始公式为:
Figure PCTCN2021129524-appb-000004
其中σ x和σ y分别是向量x和向量y的标准差,cov(x,y)是x和y向量的协方差。
根据原始公式,计算电池单体i与电池单体j之间的皮尔逊相关系数:
Figure PCTCN2021129524-appb-000005
式中,
Figure PCTCN2021129524-appb-000006
Figure PCTCN2021129524-appb-000007
分别代表在采样时间t 0到采样时间t s之间的电池单体i与电池单体j的电池单体电压。其中j按照下式定义:
Figure PCTCN2021129524-appb-000008
经过计算,得到相邻两个电池单体的皮尔森相关系数向量为:
Figure PCTCN2021129524-appb-000009
其中,图3中V 1,V 2,...V D均表示电池单体的电压值。
步骤10212:根据确定的相邻两个电池单体间的皮尔森相关系数确定电池单体的相关系数。相关系数的具体计算公式为:
Figure PCTCN2021129524-appb-000010
计算得到电池单体相关系数向量:
Figure PCTCN2021129524-appb-000011
步骤10213:获取每一电池单体的电压值以及所有电池单体的电压均值。
步骤10214:根据每一电池单体的电压值以及所有电池单体的电压均值确定电池单体的波动方差。
其中,波动方差的确定过程为:
根据每一电池单体的电压值以及所有电池单体的电压均值对电池单体进行趋势化处理后,得到趋势化向量。具体的,对于某一帧,首先求出该帧中所有电池单体的电压均值,然后该帧中每个电池单体的电压值都减去该帧所有电池单体的电压均值,记为DV。对于电池单体i,在t 0-t s时间段内,去趋势化后形成的趋势化向量表示为:
Figure PCTCN2021129524-appb-000012
根据趋势化向量确定电池单体的波动方差。具体的,电池单体i在t 0-t s时间段内的波动方差按照如下公式进行计算:
Figure PCTCN2021129524-appb-000013
其中,
Figure PCTCN2021129524-appb-000014
是向量
Figure PCTCN2021129524-appb-000015
中所有值的平均值,l表示采样点个数。最终形成电池单体波动方差矩阵:
Figure PCTCN2021129524-appb-000016
步骤10215:根据电池单体的相关系数和电池单体的波动方差构建样本集。
步骤1022:采用K-means聚类算法基于样本集将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇。进行聚类的过程具体为:
输入是样本集D={q 1,q 2,...q n},其中q i=[S i,r i],表示电池单体i的两个参数(波动方差和相关系数),默认聚类的簇分为两类,即h-2,每一簇的中心分别用c 1和c 2表示。
在所有电池单体中随机选取两个电池单体q i和q j作为簇中心。
计算所有每个电池单体样本到簇中心的欧式距离d i,每个样本按照到簇中心c 1、c 2距离最小的原则确定其所在的簇。确定原则用公式(6)表示。计算电池单体i两个参数到两个簇中心的欧式距离,如公式(7)所示,此时,若距c 1的欧式距离小于距c 2的欧式距离,则将电池单体i归于以c 1为簇中心的簇,反之归于以c 2为簇中心的簇。
c (h):=arg min h||q (i)-u h|| 2 h=1,2    (6)
采用如下公式重新计算簇中心:
Figure PCTCN2021129524-appb-000017
其中,公式(7)中分子代表每一簇中所有值的和,分母代表每一簇的样本数量。
更换簇中心,直到簇中心不再改变为止,得到每簇的中心为c 1和c 2,其中
Figure PCTCN2021129524-appb-000018
步骤103:确定异常电池单体簇和正常电池单体簇中电池单体的数量比,并分别确定异常电池单体簇中簇中心和正常电池单体簇中簇中心的相关参数。相关参数包括:相关系数和波动方差。
其中,正常电池单体簇中电池单体的数量比采用以下公式确定:
Figure PCTCN2021129524-appb-000019
式中,k a为数量比,n abnormal与n normal分别代表异常单体簇和正常单体簇中电池单体的数量。
步骤104:根据相关参数确定异常电池单体簇的簇中心与正常电池单体簇的簇中心间的欧式距离。计算欧式距离的公式如下:
Figure PCTCN2021129524-appb-000020
步骤105:获取预设阈值。预设阈值包括:数量比阈值k s和欧式距离阈值d s
步骤106:根据数量比与数量比阈值间的关系,以及欧式距离与欧式距离阈值间的关系确定异常电池单体,并输出异常电池单体的序号。具体的,如果d>d s且k a<k s,则将h设置为2,并且报警异常簇中的电池单体编号。否则,将h置为1,不给出异常单体报警。本发明优选的k s=5%,d s=3,但不限于此。
此外,对应于上述提供的基于聚类分析的电池***在线故障诊断方法,本发明还提供了一种基于聚类分析的电池***在线故障诊断***。如图4所示,该基于聚类分析的电池***在线故障诊断***,包括:运行数据获取模块1、电压矩阵形成模块2、簇分类模块3、参数确定模块4、欧式距离确定模块5、阈值获取模块6和异常电池单体确定模块7。
其中,运行数据获取模块1用于获取电动汽车的运行数据。运行数据包括:每一电池单体的电压、电流和温度。
电压矩阵形成模块2用于根据运行数据形成电压矩阵。电压矩阵的行代表电池单体序号,电压矩阵的列代表时间序列。
簇分类模块3用于采用K-means聚类算法,根据电压矩阵将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇。
参数确定模块4用于确定异常电池单体簇和正常电池单体簇中电池单体的数量比,并分别确定异常电池单体簇中簇中心和正常电池单体簇中簇中心的相关参数。相关参数包括:相关系数和波动方差。
欧式距离确定模块5用于根据相关参数确定异常电池单体簇的簇中心与正常电池单体簇的簇中心间的欧式距离。
阈值获取模块6用于获取预设阈值。预设阈值包括:数量比阈值和欧式距离阈值。
异常电池单体确定模块7用于根据数量比与数量比阈值间的关系,以及欧式距离与欧式距离阈值间的关系确定异常电池单体,并输出异常电池单体的序号。
优选地,上述簇分类模块3具体包括:样本集构建子模块和簇分类子模块。
其中,样本集构建子模块用于根据电压矩阵构建样本集。样本集包括:多个由每一电池单体的相关系数和每一电池单体波动方差构成的元素。
簇分类子模块用于采用K-means聚类算法基于样本集将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇。
优选地,为了进一步提高电池单体确定的准确性,上述样本集构建子模块可以包括:皮尔森相关系数确定单元、相关系数确定单元、电压值获取单元、波动方差确定单元和样本集构建单元。
其中,皮尔森相关系数确定单元用于确定电压矩阵中相邻两个电池单体间的皮尔森相关系数。
相关系数确定单元用于根据确定的相邻两个电池单体间的皮尔森相关系数确定电池单体的相关系数。
电压值获取单元用于获取每一电池单体的电压值以及所有电池单体的电压均值。
波动方差确定单元用于根据每一电池单体的电压值以及所有电池单体的电压均值确定电池单体的波动方差。
样本集构建单元用于根据电池单体的相关系数和电池单体的波动方差构建样本集。
优选地,上述波动方差确定单元具体包括:趋势化向量确定子单元和波动方差确定子单元。
其中,趋势化向量确定子单元用于根据每一电池单体的电压值以及所有电池单体的电压均值对电池单体进行趋势化处理后,得到趋势化向量。
波动方差确定子单元用于根据趋势化向量确定电池单体的波动方差。
本说明书中各个实施例采用递进的方式描述,每个实施例重点说明的都是与其他实施例的不同之处,各个实施例之间相同相似部分互相参见即可。对于实施例公开的***而言,由于其与实施例公开的方法相对应,所以描述的比较简单,相关之处参见方法部分说明即可。
本文中应用了具体个例对本发明的原理及实施方式进行了阐述,以上实施例的说明只是用于帮助理解本发明的方法及其核心思想;同时,对于本领域的一般技术人员,依据本发明的思想,在具体实施方式及应用范围上均会有改变之处。综上所述,本说明书内容不应理解为对本发明的限制。

Claims (8)

  1. 一种基于聚类分析的电池***在线故障诊断方法,其特征在于,包括:
    获取电动汽车的运行数据;所述运行数据包括:每一电池单体的电压、电流和温度;
    根据所述运行数据形成电压矩阵;所述电压矩阵的行代表电池单体序号,所述电压矩阵的列代表时间序列;
    采用K-means聚类算法,根据所述电压矩阵将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇;
    确定所述异常电池单体簇和所述正常电池单体簇中电池单体的数量比,并分别确定所述异常电池单体簇中簇中心和所述正常电池单体簇中簇中心的相关参数;所述相关参数包括:相关系数和波动方差;
    根据所述相关参数确定所述异常电池单体簇的簇中心与所述正常电池单体簇的簇中心间的欧式距离;
    获取预设阈值;所述预设阈值包括:数量比阈值和欧式距离阈值;
    根据所述数量比与所述数量比阈值间的关系,以及所述欧式距离与所述欧式距离阈值间的关系确定异常电池单体,并输出所述异常电池单体的序号。
  2. 根据权利要求1所述的基于聚类分析的电池***在线故障诊断方法,其特征在于,所述采用K-means聚类算法,根据所述电压矩阵将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇,具体包括:
    根据所述电压矩阵构建样本集;所述样本集包括:多个由每一电池单体的相关系数和每一电池单体波动方差构成的元素;
    采用所述K-means聚类算法基于所述样本集将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇。
  3. 根据权利要求2所述的基于聚类分析的电池***在线故障诊断方法,其特征在于,所述根据所述电压矩阵构建样本集,具体包括:
    确定所述电压矩阵中相邻两个电池单体间的皮尔森相关系数;
    根据确定的相邻两个电池单体间的皮尔森相关系数确定电池单体的相关系数;
    获取每一电池单体的电压值以及所有电池单体的电压均值;
    根据所述每一电池单体的电压值以及所有电池单体的电压均值确定电池单体的波动方差;
    根据所述电池单体的相关系数和所述电池单体的波动方差构建所述样本集。
  4. 根据权利要求3所述的基于聚类分析的电池***在线故障诊断方法,其特征在于,所述根据所述电压值和电压均值确定电池单体的波动方差,具体包括:
    根据所述每一电池单体的电压值以及所有电池单体的电压均值对所述电池单体进行趋势化处理后,得到趋势化向量;
    根据所述趋势化向量确定电池单体的波动方差。
  5. 一种基于聚类分析的电池***在线故障诊断***,其特征在于,包括:
    运行数据获取模块,用于获取电动汽车的运行数据;所述运行数据包括:每一电池单体的电压、电流和温度;
    电压矩阵形成模块,用于根据所述运行数据形成电压矩阵;所述电压矩阵的行代表电池单体序号,所述电压矩阵的列代表时间序列;
    簇分类模块,用于采用K-means聚类算法,根据所述电压矩阵将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇;
    参数确定模块,用于确定所述异常电池单体簇和所述正常电池单体簇中电池单体的数量比,并分别确定所述异常电池单体簇中簇中心和所述正常电池单体簇中簇中心的相关参数;所述相关参数包括:相关系数和波动方差;
    欧式距离确定模块,用于根据所述相关参数确定所述异常电池单体簇的簇中心与所述正常电池单体簇的簇中心间的欧式距离;
    阈值获取模块,用于获取预设阈值;所述预设阈值包括:数量比阈值和欧 式距离阈值;
    异常电池单体确定模块,用于根据所述数量比与所述数量比阈值间的关系,以及所述欧式距离与所述欧式距离阈值间的关系确定异常电池单体,并输出所述异常电池单体的序号。
  6. 根据权利要求5所述的基于聚类分析的电池***在线故障诊断***,其特征在于,所述簇分类模块具体包括:
    样本集构建子模块,用于根据所述电压矩阵构建样本集;所述样本集包括:多个由每一电池单体的相关系数和每一电池单体波动方差构成的元素;
    簇分类子模块,用于采用所述K-means聚类算法基于所述样本集将电动汽车中的电池单体分为异常电池单体簇和正常电池单体簇。
  7. 根据权利要求6所述的基于聚类分析的电池***在线故障诊断***,其特征在于,所述样本集构建子模块具体包括:
    皮尔森相关系数确定单元,用于确定所述电压矩阵中相邻两个电池单体间的皮尔森相关系数;
    相关系数确定单元,用于根据确定的相邻两个电池单体间的皮尔森相关系数确定电池单体的相关系数;
    电压值获取单元,用于获取每一电池单体的电压值以及所有电池单体的电压均值;
    波动方差确定单元,用于根据所述每一电池单体的电压值以及所有电池单体的电压均值确定电池单体的波动方差;
    样本集构建单元,用于根据所述电池单体的相关系数和所述电池单体的波动方差构建所述样本集。
  8. 根据权利要求7所述的基于聚类分析的电池***在线故障诊断***,其特征在于,所述波动方差确定单元具体包括:
    趋势化向量确定子单元,用于根据所述每一电池单体的电压值以及所有电池单体的电压均值对所述电池单体进行趋势化处理后,得到趋势化向量;
    波动方差确定子单元,用于根据所述趋势化向量确定电池单体的波动方差。
PCT/CN2021/129524 2021-01-18 2021-11-09 一种基于聚类分析的电池***在线故障诊断方法和*** WO2022151819A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110059389.7 2021-01-18
CN202110059389.7A CN112858919B (zh) 2021-01-18 2021-01-18 一种基于聚类分析的电池***在线故障诊断方法和***

Publications (1)

Publication Number Publication Date
WO2022151819A1 true WO2022151819A1 (zh) 2022-07-21

Family

ID=76007154

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/129524 WO2022151819A1 (zh) 2021-01-18 2021-11-09 一种基于聚类分析的电池***在线故障诊断方法和***

Country Status (2)

Country Link
CN (1) CN112858919B (zh)
WO (1) WO2022151819A1 (zh)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115061049A (zh) * 2022-08-08 2022-09-16 山东卓朗检测股份有限公司 一种数据中心ups电池故障快速检测方法及***
CN115366683A (zh) * 2022-08-09 2022-11-22 北京理工大学 一种新能源汽车动力电池多维模型融合的故障诊断策略
CN115511013A (zh) * 2022-11-22 2022-12-23 河海大学 一种大规模储能电站异常电池识别方法、设备及存储介质
CN115840157A (zh) * 2022-12-08 2023-03-24 斯润天朗(合肥)科技有限公司 基于eof分析的锂电池电性能指标协调性分析***
CN116106758A (zh) * 2023-03-23 2023-05-12 华能新能源股份有限公司山西分公司 基于数据驱动的电池故障诊断方法及***
CN116404186A (zh) * 2023-06-08 2023-07-07 西安黄河电子技术有限公司 一种功率型锂锰电池生产***
CN116502112A (zh) * 2023-06-29 2023-07-28 深圳市联明电源有限公司 一种新能源电源测试数据管理方法及***
CN116756595A (zh) * 2023-08-23 2023-09-15 深圳市森瑞普电子有限公司 一种导电滑环故障数据采集监测方法
CN116749812A (zh) * 2023-08-14 2023-09-15 南通百仕灵新能源科技有限公司 一种新能源充电设备的自适应充电方法
CN117648589A (zh) * 2024-01-30 2024-03-05 云储新能源科技有限公司 一种储能电池热失控预警方法、***、电子设备及介质
CN117665629A (zh) * 2024-01-31 2024-03-08 成都深瑞同华科技有限公司 储能电池仓的电压一致性评估方法、装置及存储介质
CN117892095A (zh) * 2024-03-14 2024-04-16 山东泰开电力电子有限公司 一种储能***用的散热***故障智能检测方法
CN118094449A (zh) * 2024-04-26 2024-05-28 山东瑞福锂业有限公司 基于工业互联网的生产智能监控方法、***及装置
CN118249467A (zh) * 2024-05-20 2024-06-25 深圳市芯科云科技有限公司 基于温度控制的智能手表无线充电方法

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112858919B (zh) * 2021-01-18 2022-04-01 北京理工大学 一种基于聚类分析的电池***在线故障诊断方法和***
CN113820333B (zh) * 2021-09-16 2024-06-07 无锡先导智能装备股份有限公司 电池极片异常检测方法、装置、上位机及检测***
CN114264957B (zh) * 2021-12-02 2024-05-07 东软集团股份有限公司 一种异常单体检测方法及其相关设备
CN114114039B (zh) * 2021-12-06 2023-10-03 湖北亿纬动力有限公司 一种电池***的单体电芯一致性的评估方法和装置
CN114839541A (zh) * 2022-05-18 2022-08-02 山东大学 基于k均值聚类的动力电池组不一致性诊断方法及***
CN115877222A (zh) * 2023-02-14 2023-03-31 国网浙江省电力有限公司宁波供电公司 一种储能电站故障检测方法、装置、介质及储能电站
CN116482560B (zh) * 2023-06-21 2023-09-12 中国华能集团清洁能源技术研究院有限公司 电池故障检测方法、装置、电子设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106371021A (zh) * 2015-07-21 2017-02-01 三星电子株式会社 用于估计电池的状态的方法和设备
CN108254689A (zh) * 2016-12-29 2018-07-06 中国电信股份有限公司 电池组反极单体电池检测方法和***
CN111929591A (zh) * 2020-08-21 2020-11-13 彩虹无线(北京)新技术有限公司 故障电池检测方法、装置、设备及计算机存储介质
CN112858919A (zh) * 2021-01-18 2021-05-28 北京理工大学 一种基于聚类分析的电池***在线故障诊断方法和***

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4370409B2 (ja) * 2005-06-22 2009-11-25 国立大学法人東北大学 がんの予後予測方法
CN110398382A (zh) * 2019-06-28 2019-11-01 南京康尼机电股份有限公司 一种基于dpc的地铁车门***性能衰退检测方法
CN110794305B (zh) * 2019-10-14 2021-03-16 北京理工大学 一种动力电池故障诊断方法和***
CN111090050A (zh) * 2020-01-21 2020-05-01 合肥工业大学 一种基于支持向量机和k均值的锂电池故障诊断方法
CN111707951B (zh) * 2020-06-22 2021-04-06 北京理工大学 一种电池组一致性评估方法及***

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106371021A (zh) * 2015-07-21 2017-02-01 三星电子株式会社 用于估计电池的状态的方法和设备
CN108254689A (zh) * 2016-12-29 2018-07-06 中国电信股份有限公司 电池组反极单体电池检测方法和***
CN111929591A (zh) * 2020-08-21 2020-11-13 彩虹无线(北京)新技术有限公司 故障电池检测方法、装置、设备及计算机存储介质
CN112858919A (zh) * 2021-01-18 2021-05-28 北京理工大学 一种基于聚类分析的电池***在线故障诊断方法和***

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115061049B (zh) * 2022-08-08 2022-11-01 山东卓朗检测股份有限公司 一种数据中心ups电池故障快速检测方法及***
CN115061049A (zh) * 2022-08-08 2022-09-16 山东卓朗检测股份有限公司 一种数据中心ups电池故障快速检测方法及***
CN115366683B (zh) * 2022-08-09 2024-05-24 北京理工大学 一种新能源汽车动力电池多维模型融合的故障诊断策略
CN115366683A (zh) * 2022-08-09 2022-11-22 北京理工大学 一种新能源汽车动力电池多维模型融合的故障诊断策略
CN115511013A (zh) * 2022-11-22 2022-12-23 河海大学 一种大规模储能电站异常电池识别方法、设备及存储介质
CN115840157B (zh) * 2022-12-08 2023-08-22 斯润天朗(合肥)科技有限公司 基于eof分析的锂电池电性能指标协调性分析***
CN115840157A (zh) * 2022-12-08 2023-03-24 斯润天朗(合肥)科技有限公司 基于eof分析的锂电池电性能指标协调性分析***
CN116106758A (zh) * 2023-03-23 2023-05-12 华能新能源股份有限公司山西分公司 基于数据驱动的电池故障诊断方法及***
CN116106758B (zh) * 2023-03-23 2024-01-30 华能新能源股份有限公司山西分公司 基于数据驱动的电池故障诊断方法及***
CN116404186A (zh) * 2023-06-08 2023-07-07 西安黄河电子技术有限公司 一种功率型锂锰电池生产***
CN116404186B (zh) * 2023-06-08 2023-09-19 西安黄河电子技术有限公司 一种功率型锂锰电池生产***
CN116502112A (zh) * 2023-06-29 2023-07-28 深圳市联明电源有限公司 一种新能源电源测试数据管理方法及***
CN116502112B (zh) * 2023-06-29 2023-10-24 深圳市联明电源有限公司 一种新能源电源测试数据管理方法及***
CN116749812B (zh) * 2023-08-14 2023-10-20 南通百仕灵新能源科技有限公司 一种新能源充电设备的自适应充电方法
CN116749812A (zh) * 2023-08-14 2023-09-15 南通百仕灵新能源科技有限公司 一种新能源充电设备的自适应充电方法
CN116756595B (zh) * 2023-08-23 2023-12-01 深圳市森瑞普电子有限公司 一种导电滑环故障数据采集监测方法
CN116756595A (zh) * 2023-08-23 2023-09-15 深圳市森瑞普电子有限公司 一种导电滑环故障数据采集监测方法
CN117648589A (zh) * 2024-01-30 2024-03-05 云储新能源科技有限公司 一种储能电池热失控预警方法、***、电子设备及介质
CN117648589B (zh) * 2024-01-30 2024-05-14 云储新能源科技有限公司 一种储能电池热失控预警方法、***、电子设备及介质
CN117665629A (zh) * 2024-01-31 2024-03-08 成都深瑞同华科技有限公司 储能电池仓的电压一致性评估方法、装置及存储介质
CN117892095A (zh) * 2024-03-14 2024-04-16 山东泰开电力电子有限公司 一种储能***用的散热***故障智能检测方法
CN117892095B (zh) * 2024-03-14 2024-05-28 山东泰开电力电子有限公司 一种储能***用的散热***故障智能检测方法
CN118094449A (zh) * 2024-04-26 2024-05-28 山东瑞福锂业有限公司 基于工业互联网的生产智能监控方法、***及装置
CN118249467A (zh) * 2024-05-20 2024-06-25 深圳市芯科云科技有限公司 基于温度控制的智能手表无线充电方法

Also Published As

Publication number Publication date
CN112858919A (zh) 2021-05-28
CN112858919B (zh) 2022-04-01

Similar Documents

Publication Publication Date Title
WO2022151819A1 (zh) 一种基于聚类分析的电池***在线故障诊断方法和***
CN107957562B (zh) 一种锂离子电池剩余寿命在线预测方法
WO2020077682A1 (zh) 一种服务质量评估模型的训练方法及装置
CN107422266B (zh) 一种大容量电池储能***的故障诊断方法及装置
CN104133143B (zh) 一种基于Hadoop云计算平台的电网线路故障诊断***及方法
CN114114039B (zh) 一种电池***的单体电芯一致性的评估方法和装置
CN113030763A (zh) 一种车用动力电池一致性差异的潜在风险诊断***
CN114386537B (zh) 基于CatBoost的锂电池故障诊断方法、装置及电子设备
CN105866725A (zh) 一种基于聚类分析和云模型的智能电表故障分类方法
CN112816881B (zh) 电池压差异常检测方法、装置及计算机存储介质
CN112485747A (zh) 一种电容式电压互感器误差状态评估方法
CN116609676B (zh) 一种基于大数据处理的混合储能电池状态监控方法及***
CN112904219A (zh) 一种基于大数据的动力电池健康状态的预测方法
CN114801751A (zh) 一种基于数据分析的汽车电池故障预测***
WO2023226355A1 (zh) 基于多源感知的双向锂离子电池故障检测方法及***
CN111527692A (zh) 判定装置、太阳光发电***、判定方法及判定程序
Zhang et al. Fault diagnosis of real-scenario battery systems based on modified entropy algorithms in electric vehicles
CN116381494A (zh) 电池最大可用容量预测方法、装置、电子设备及存储介质
CN116482540A (zh) 电池电压不一致性的分析预测方法、装置及***
CN114579659A (zh) 一种梯次利用动力电池利用潜力评估及分选***和方法
CN114280494A (zh) 一种基于云端协同的锂电池健康状态监测方法及装置
CN107730148B (zh) 一种输电线路隐患预警方法和***
CN113379005A (zh) 一种电网电力设备能源智能管理***及方法
CN111273126A (zh) 一种配电网拓扑快速感知方法
CN116842464A (zh) 一种电池***soc估计方法

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: 21919010

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21919010

Country of ref document: EP

Kind code of ref document: A1