Fault Diagnosis Method of DC Microgrid Multi-Port Power Electronic Circuits Based on WOA-KELM Algorithm

The faults in DC microgrid multi-port power electronic circuits are complex and diverse. In order to better adapt to different types of fault data and complex circuit environment, and improve the generalization ability and accuracy of fault diagnosis, a fault diagnosis method for DC microgrid multi-port power electronic circuits based on WOA-KELM algorithm is studied. Through the wavelet packet analysis algorithm, the wavelet coefficient energy of DC microgrid multi-port power electronic circuit signal is extracted as a fault feature, and it is encoded to speed up fault diagnosis efficiency; the WOA algorithm is used to optimize the penalty parameters of KELM algorithm to improve the accuracy of fault diagnosis; in the optimized KELM algorithm, the coded fault feature samples are input, and the fault diagnosis results of DC microgrid multi-port power electronic circuits are output. According to the powerful nonlinear mapping ability of the kernel function, the generalization ability of the KELM algorithm is enhanced, so that it can better adapt to different types of fault data and complex circuit environment. Experiments show that this method can effectively extract fault features of multi-port power electronic circuits and complete feature coding; this method can effectively diagnose the faults of DC microgrid multi-port power electronic circuits; in different complex circuit scenarios, the fault reliability of this method for diagnosing power circuit faults is higher than 0.6, that is, the fault diagnosis accuracy is high.