Fault diagnosis method of HV circuit breaker based on wavelet packet time-frequency entropy and BP neural network

Abstract

High-voltage High-voltage circuit breakers are the most important control and protection equipment in power systems and their reliable operation is critical to power systems. However, the mechanical failure of high-voltage circuit breakers occurs frequently. The vibration signals of high-voltage circuit breakers contain abundant fault information. The change of vibration signals reflects the mechanical state of the circuit breakers. The extraction and classification of vibration signals are very important for fault diagnosis of HV circuit Breaker. In this paper, the packet time-frequency entropy is used to extract the characteristic of the vibration signal of the circuit breaker and BP neural network is used to identify the various types of fault vibration signals. Specially, the vibration signal is decomposed by wavelet packet, then construct the time-frequency entropy of the vibration signal, which is used to the feature vector of fault vibration signals. Finally, we use the BP neural network to judge the working state and fault type of the circuit breaker. The experimental results show that the combination of wavelet packet time-frequency entropy and BP neural network can effectively judge the mechanical failure of the circuit breaker.