A novel online diagnosis method for rotor eccentricity faults in squirrel-cage induction motors

Motor current signature analysis has become a widely used technique in the fault diagnosis of induction motors. Conventional rotor eccentricity fault diagnosis methods rely on fast Fourier transform (FFT) algorithms to analyze stator current spectra and identify characteristic fault frequencies. However, inherent limitations of FFT, including spectral leakage and the picket-fence effect, significantly compromise diagnostic accuracy. This study proposes a novel method employing Hilbert transform to process time-domain stator current signals, generating corresponding analytic functions. Theoretical analysis shows that the amplitude of the analysis function contains both a direct current component and an alternating current component, where the former corresponds to the fundamental frequency component in the original stator current, whereas the latter corresponds to the fault characteristic component in the original stator current during eccentricity faults. Therefore, spectral analysis of the amplitude of this analytic function can significantly enhance the diagnosis of eccentricity faults. Signal testing based on Matlab has validated the effectiveness of this method.