基于多算法融合的锂离子电池故障诊断方法

Fault Diagnosis Method for Lithium-ion Battery Based on Multi-algorithm Fusion

锂离子电池的安全和热失控问题,限制了其在各领域的广泛应用。为了更准确地检测电池异常情况中的内短路与开路故障,提出一种基于多算法融合的锂离子电池故障诊断方法。首先,采用逐次变分模态分解(sequential variational mode decomposition,SVMD)计算每个电池单体的最大电压波动,并通过箱线图和离群点检测来识别异常单体。其次,根据SVMD得到的电压均值差的绝对值,结合斜率计算和统计分析,检测潜在的内短路或开路故障。最后,利用实际锂离子电池的热失控数据对所提方法进行验证。结果表明,该方法可以在热失控发生前55 min进行预警,并能精确识别内短路或开路故障单体,具有较高的可靠性和工程应用价值。

The safety and thermal runaway problems of the lithium-ion battery limit its wide range of applications in various fields.In order to more accurately detect internal short-circuit and open-circuit faults in abnormal battery conditions,this paper proposes a multi-algorithm fusion-based fault diagnosis method for the lithium-ion battery.Firstly,the sequential variational mode decomposition(SVMD)method is used to calculate the maximum voltage fluctuation of each battery cell,and the anomalous cells are identified by box-and-line diagram and outlier detection.Secondly,the absolute value of the voltage mean difference obtained from SVMD decomposition is used to detect potential internal short or open circuit faults by combining slope calculation and statistical analysis.Finally,the paper verifies this method using thermal runaway data from actual lithium-ion batteries.The results show that the method can provide early warning 55 minutes before thermal runaway occurs and accurately identify internal short-circuit or open-circuit faulty monoliths,which is highly reliable and has high value for engineering applications.