Research on Fault Diagnosis Technology of AC Medium Voltage Vacuum Circuit Breaker Based on Wavelet Packet

Abstract

Since mechanical failure is one of the main failures of the circuit breaker, on-line monitoring of mechanical condition has important practical significance. Mechanical vibration can reflect the mechanical state of circuit breaker in the whole time series and spectrum range. Besides, the vibration signals produced by the same type of circuit breaker are similar, and the operation status of the equipment can be diagnosed and studied by the vibration signals. Therefore, the vibration signal can be used to extract features to diagnose the running state of the equipment. In this paper, for a 12 kV AC medium volt-age vacuum circuit breaker, a variety of fault conditions are designed and the vibration signal of the breaker under normal and fault conditions is collected. The original signals are decomposed by wavelet packet and processed by method of band energy, energy entropy and Euclidean distance respectively, and then the eigenvectors including vibration characteristics are obtained. Finally Support Vector Machine is used to distinguish the normal and fault states, which provides a basis for the realization of on-line fault diagnosis technology for circuit breakers.