The invention provides a vehicle power battery system fault diagnosis method based on an LSTM-GAN, and the method comprises the steps: carrying out the unsupervised learning training of a relation between the voltage of each single battery and each characterization parameter through the historical operation data of a vehicle based on an LSTM-GAN model; and carrying out statistical analysis on a deviation coefficient obtained by the LSTM-GAN model of a vehicle using the same battery system to realize threshold detection of single voltage abnormity. According to the method, the problem that the state of the single battery in the power battery system is difficult to monitor is effectively solved, and the dependence on the data quantity and quality is small, so that the calculation overhead cost can be reduced while the accuracy and efficiency of fault diagnosis can be improved.
本发明提供了一种基于LSTM‑GAN的车用动力电池系统故障诊断方法,利用车辆历史运行数据并基于LSTM‑GAN的模型对各电池单体电压与各表征参数间关系进行无监督学习训练,并通过对使用相同电池系统车辆由LSTM‑GAN模型得到的偏差系数进行统计分析,来实现单体电压异常的阈值检测。该方法有效解决了动力电池系统中电池单体状态监测困难的问题,对数据数量与质量的依赖性较小,故能够在提升故障诊断的精确性及效率的同时降低计算开销成本。
Vehicle power battery system fault diagnosis method based on LSTM-GAN
一种基于LSTM-GAN的车用动力电池系统故障诊断方法
2023-11-24
Patent
Electronic Resource
Chinese
Turnout fault diagnosis method and system based on 1DCNN-LSTM
European Patent Office | 2023
|Power battery system fault diagnosis method, diagnosis system, design method and vehicle
European Patent Office | 2023
|Bearing Fault Diagnosis Method of Bearing Based on LSTM Auto-Encoder
British Library Conference Proceedings | 2022
|