Bearing damage detection based on statistical discrimination of stator current

In this study, a method is presented for discriminating stator current signals from two classes, motors in normal condition and ones with a bearing failure. The study has two main objectives, to verify the former results on the bearing fault detection based on the stator current, and to show the applicability of the proposed method for the diagnosis based on frequency content. The method is based on statistical analysis of Gabor filter responses. In the empirical part of this study where the method is applied to the real measurements, the results indicate that it is not possible to detect a bearing fault if the internal radial clearance of the bearing is not adequate.