Induction Machine Bearing Fault Detection Using Empirical Wavelet Transform

The detection of faults related to the optimal condition induction motors is an important task avoid malfunction or loss motor, thus avoiding high repair replacement costs and in efficiency process which they belong. These are not limited a single area; mechanical electrical problems can cause fault. Specifically, bearing motor subjected several effects that faults, significant breakdowns machinery. This article proposes methodology for detecting on motor. first part uses signal processing method called empirical wavelet transform (EWT), decomposes vibration into multiple components extract series amplitude frequency modulated (AM-FM) with Fourier spectrum. First, data collected normal operating other damage due perforation. Then, three types goodness-of-fit tests used, Kuiper, Kolmogorov–Smirnov, Pearson chi-square, classify signals determine ones belong damaged engine. Finally, experimental results show EWT conjunction proposed goodness achieves competitive precision diagnosing motor-bearing faults.