8th International Conference on Advanced Power System Automation and Protection (APAP2019)

868 / 2019-04-30 10:14:29

A Survey of Artificial Intelligent Fault Diagnosis in Power Electronics

supervised learning,Unsupervised learning,signal processing,Fault Diagnosis

With continuous increase in complexity of power electronic system, there is less stability for power electronic system, productivity decrease, which greatly necessitates to detect and identify any types of faults as soon as possible. Artificial intelligence algorithms and signal based methods have been made great breakthroughs in feature engineering and pattern recognition. Therefore, the advantages of using these algorithms in fault diagnosis of power electronic systems are enormous. In this paper, the characteristics of several fault diagnosis algorithms are analyzed in detail, and some new research directions are proposed according to the problems existing in power electronic system fault diagnosis methods.