Research on combined diagnosis of mechanical fault vibration-sound signal of high voltage circuit breaker based on EEMD-energy entropy feature

Abstract

Current research reveal many problems in the process of mechanical fault diagnosis of high-voltage circuit breakers (HVCBs). One of the problems concerns the vibration signals of HVCBs involving a wide range and large amplitude. This makes it difficult to monitor. A single vibration signal is seriously affected by the position. And it is easy to saturate and top off, which can’t fully display the mechanical fault information of HVCBs.Therefore, this paper collects vibration and sound signals for the typical five working conditions of the HVCBs, and proposes an EEMD-energy entropy feature extraction method, then uses the KNN algorithm to diagnose the five working conditions. Compared with the diagnosis results under a single vibration signal and a single sound signal, the superiority of the vibration-sound signal combined analysis and the effectiveness of the proposed feature extraction method areverified, which provides a new idea for the study of mechanical fault diagnosis of HVCBs.