On-line rotor fault detection in squirrel cage induction motors based on wavelet analysis

As the amplitude of the rotor fault feature components are always submerged in the fundamental components, they are difficult to be detected by the conventional Fourier transform, this paper takes advantage of the distinguishing feature of the wavelet transform in discerning mutable signal to extract them. Firstly, Morlet wavelet is chosen as the mother wavelet, and the current signal of the stator is analyzed by MRA, subsequently, with the reconstruction of the selected sub-spaces, the motor´s fault feature components are extracted successfully, which provides an effective way for the on-line diagnosis of the induction motors.