Adaptive Weighted VMD-WPEE Method of Power-Electronic-Circuit Multiple-Parameter-Fault Diagnosis

To improve the diagnostic accuracy rate of multiple parameter faults in widely used power electronic circuits, solving for the overlapping parameter fault characteristic signal is crucial. This paper proposes a novel approach for diagnosing the multiple parameter faults of power electronic circuits that can adaptively adjust the weights of each fault feature parameter according to the diagnostic accuracy rate to enhance their discriminability. According to the properties of multiple parameter faults in power electronic circuits, the measured signal that has been decomposed via the variational mode decomposition and wavelet packet energy entropy (VMD-WPEE) is used to extract the feature parameters. Furthermore, a classifier is selected and the threshold of the diagnostic accuracy rate is set. Then, the weight of each fault feature parameter is adjusted until the classifier yields the desired results. Finally, the proposed approach is applied to other classifiers and feature extraction methods. The high accuracy, high robustness, and advantages of the multiple-parameter-fault diagnosis method are demonstrated by the simulation and experimental results.