Detection of Shaft Misalignment of a PMSM using Zoom-FFT

In this paper, Shaft Misalignment of a Permanent Magnet Synchronous Motor (PMSM) is investigated, through a computationally-efficient method, Zoom-FFT. Conventional FFT computes a wide-range frequency spectrum while only a part of the frequency spectrum is necessary for fault detection. As conventional Zoom-FFT may lead to higher complexity for narrow fault-related frequency bands, a Two Stage Zoom-FFT analysis has been proposed, which offers a low-complexity solution, taking advantage of the current measurement sensors integrated in the PMSM drive system. The proposed method has been validated through experimental testing under different operating conditions.