Enhanced Diagnosis of the Broken Bar Fault in Induction Motors Through the Wiener Filter

For better diagnosis of the broken bar fault (BBF) of the induction motor, the performance of the Wiener filter (WF) for suppressing the fundamental frequency component (FFC) of the stator current signal (SCS) is explored through simulation and experiments. The WF utilizes the FFC of the supply voltage signal (SVS) to estimate the FFC of the SCS. Due to orthogonality of sinusoidal signals, the estimate of the FFC of the SCS contains little information of the sideband harmonics (SBHs). Then, subtracting the estimate of the FFC from the SCS suppresses the FFC significantly, with little impact on the SBHs, especially their amplitudes, which is validated by simulation. Thus, as revealed by the experimental results, the SBHs with their amplitudes can be unambiguously identified from the spectrum of the SCS processed by the WF, leading to feasibility of diagnosis of the BBF at its incipient stage through monitoring the amplitudes of the SBHs. Comparison of the performance of FFC suppression of the WF with that of the Kalman filter further illustrates superiority of the WF for FFC suppression. Therefore, our research demonstrates the aforementioned WF is pretty effective in the situation where diagnosis of induction motor faults through motor current signature analysis is hampered by the widened peak of the FFC of the SCS due to spectral leakage, especially in the case of the BBF.