Induction Motor Fault Diagnosis Using Discrete Wavelet Transform

Abstract

In this paper, a general methodology is constructed by using experimentally measured stator current signals under full load condition of an induction motor. The measured stator current data for various single-and multi-electrical faults of the induction motor are investigated to obtain signatures for fault diagnosis. In this study, the discrete wavelet transform (DWT) is chosen for signal processing. The threshold and energy values at each decomposition level for the DWT analysis are evaluated. The threshold values appear to be more consistent than the energy values at different data windows of the measurement data, and thus, the threshold can serve as a reliable parameter for fault diagnosis.