The method of machinery fault diagnosis for circuit breaker based on bispectrum and Hilbert-Huang transform

During the circuit breaker operating process, the of acoustic signal changing contains its machinery state information. Circuit breaker has complex structure, the fault occurrence and development lead to acoustic signal having a unsteady state of the complex characteristics. The paper puts forward a circuit breaker state diagnosis method based on bispectrum and Hilbert-Huang transform. Basing on the bispectrum analysis of the normal and abnormal closed acoustic signal, it is found that the high frequency component´s energy distribution of acoustic signal can reflect the work state of circuit breaker; according to energy distribution characteristics of high frequency component through empirical mode decomposition, a method of extracting fault feature is proposed; simulation experiments take energy entropy of 500 ~ 3000 Hz intrinsic mode functions as the basis of circuit breaker´s fault diagnosis, which can purify the influence generated by environment of low frequency on the result of the analysis.