Research on Fault Diagnosis and Application of Battery Systems Based on Multi-entropy Fusion

Efficient, stable, and reliable battery fault diagnosis is crucial to ensure the safe operation of new energy vehicles. Firstly,based on the eigenanalysis of the screened entropy, it is divided into temporal entropy and multiscale entropy, and five typical entropies are selected to carry out the validation of power battery fault diagnosis effect. The results show that temporal entropy is fast and low in computation, which is suitable for real-time fault diagnosis, while multiscale entropy is more prominent for the diagnosis ability of fluctuating abnormalities. The results show that temporal entropy is fast and low computational power, which is suitable for real-time fault diagnosis, while multi-scale entropy is more prominent for the diagnosis of fluctuating abnormalities. Then the two entropies are selected separately to further explore the intrinsic relationship between key computational factors and fault diagnosis effect. The results show that the evolution law of logarithmic function is shown between the diagnosis effect of modified Shannon entropy and computational window, and the relationship between the diagnosis effect of modified multiscale entropy and scale factor is approximately normal distribution. Finally, a multi-entropy fusion fault diagnosis strategy for practical engineering applications is proposed for the comprehensive diagnosis of two fault types, namely, size class and fluctuation class, and specific application ideas are given in three aspects, namely, online state identification, real-time fault diagnosis and comprehensive fault troubleshooting. The above research results can significantly improve the efficiency and diagnostic coverage of power battery fault diagnosis, realize high-efficiency power battery online condition monitoring and comprehensive fault detection, and have important theoretical guidance significance and broad application prospects for the subsequent development of high-safety power battery management system and health supervision system.