Frequency Extraction of Current Signal Spectral Components: A New Tool for the Detection of Rotor Electrical Faults in Induction Motors

Abstract

This work expands the classical current signature analysis in induction machines in a two-stage spectral decomposition manner. The proposed methodology can be summarized in two main steps: initially, the current signals are analyzed using a time frequency representation, with the analysis focusing on the steady-state regime; thereafter, frequency extraction is applied to the spectral signatures of interest, aiming to identify specific fault related harmonic subcomponents induced by the fault related speed ripple effect. The proposed approach is verified experimentally on a 4 kW induction motor.