ICA and wavelet packet decomposition approaches for monitoring the incipient bearing damage in electrical motors

Abstract

Contemporary engineering approach shows a trend in which the predictive maintenance i.e., online monitoring and early detection of the failures take the place of the periodic maintenance. This study focused on capturing an aging trend of the bearing of a 5-HP induction motor. Independent Components Analysis (ICA) based on the wavelet package analysis has been used to foreseen the bearing damage for predictive maintenance information.