THE APPLICATION OF WAVELET ANALYSIS TO DETECT ROTOR BAR BREAKING FAULT OF SQUIRREL CAGE ASYNCHRONOUS MOTOR ON-LINE
Liling Sun · Proceedings of the CSEE · 2001
Rotor bar breaking is one frequent fault of squirrel cage asynchronous motor. This paper briefly reviews several schemes to detect rotor bar breaking on-line, and points out that how to distinguish load oscillation and rotor bar b reaking correctly is still the essential problem up to now. Based on wavelet ana lysis, a method to determine whether the load oscillates and then locates it exa ctly is presented in this paper. In addition, the method can reflect the range o f load oscillation. At the same time, this paper discusses the feasibility using wavelet analysis to extract the feature component of rotor bar breaking fault, and draws a conclusion that wavelet analysis is inapplicable to some extent. Con sequently, on-line detection schemes must rely on Fourier transform to extract t he feature component. Based on the researches above mentioned, this paper presen ts a new on-line detection scheme that uses wavelet analysis, self-adaptive fi lt er and continuous subdivision Fourier transform synthetically. Results of numeri cal simulation and experiments show that the new scheme is valid and effective.