A comparative study of induction motor current signature analysis techniques for mechanical faults detection

Abstract

This paper provides a comprehensive comparison of motor current signature analysis techniques for broken rotor bars and air gap eccentricity detection in induction motors. Four characteristic categories of processing methods are investigated, which are Fourier based analysis of the phase currents, parametric and eigenanalysis methods for spectrum estimation of the phase currents, wavelet analysis of the phase currents and current space vector analysis. The analysis is at first performed through simulation of a faulty asynchronous machine in Matlab-Simulink. Subsequent to the simulation results analysis, an experimental fault test rig was designed and implemented for the investigation of the aforementioned malfunctions. In this preliminary investigation, each analysis technique along with its advantages and disadvantages is presented and briefly discussed.