Induction Motor Bearing Fault Detection with Non-stationary Signal Analysis

The purpose of this research is to identify bearing fault features. This approach uses continuous wavelet transforms as a non-stationary signal preprocessor and the singular value decomposition (SVD) technique as salient feature extraction. Simulations of a model for bearing inner race defect as well as actual bearing vibration data from a normal bearing and the defective inner race bearing are used to demonstrate the proposed method for bearing fault detection and diagnosis. The results obtained have shown that this approach is effective for bearing fault detection and diagnosis.