Volume 48

Diagnosis of Vienna Rectifier Faults Based on EMD Feature Extraction and Optimized RF Classification Chen Yang, Jianzhou Zhang, Haiqi Li

Abstract

The front-end part of the DC charging module for electric vehicles commonly utilizes the Vienna rectifier, whose stable operation directly impacts the overall status of the charging module. Therefore, focusing on the characteristics of open-circuit faults in core components such as power switches and electrolytic capacitors of the Vienna rectifier, this paper proposes a diagnostic method based on Empirical Mode Decomposition (EMD) and Whale Optimization Algorithm (WOA) optimized Random Forest (RF) algorithm. Firstly, by constructing a simulation model of the Vienna rectifier, the waveform characteristics of the input current during open-circuit faults are summarized. The fault current signal is decomposed, and feature vectors are constructed using the EMD method. These feature vectors are then input into the classification model with optimized parameters using the WOAoptimizedRandom Forest. Simulation results demonstrate that this method achieves a high fault diagnosis rate and reduces diagnosis time, providing practical guidance for fault diagnosis in DC charging piles for automobiles.

Keywords Vienna rectifier, Empirical Mode Decomposition, Whale Optimization Algorithm, Random Forest, Fault diagnosis