Experimental diagnosis of inter-turns stator fault and unbalanced voltage supply in induction motor using MCSA and DWER

Abstract

This paper presents a comparative study between two techniques of signal processing to diagnose both faults the inter-turn short circuit (ITSC) in stator windings and the unbalanced voltage supply (UVS) in induction motors. The first is considered a classical technique called Motor Current Signature Analysis (MCSA) which is based on the processing of the stator current by the Fast Fourier Transform (FFT). The second is anadvanced technique based on a Discrete Wavelet Energy Ratio (DWER) of three stator currents. The aim objective of this paper is to compare the ability and effectiveness of both techniques to detect the ITSC fault and the UVS in induction motors, and distinguishing between them. An experimental implementation tests the two diagnosis techniques.The results obtained show that the MCAS technique by the FFT analysis has a difficult to discriminate between the current harmonics due to the provide voltage unbalance and those originated by ITSC faults. Unlike the DWERtechnique, which has high sensitivity and exceptional ability to detect and distinguish between the two faults that lead to the reliability of the diagnosis system. To demonstrate that the DWER is an accurate and robust diagnosis approach are used the neural network (NN) as a tool to classify the faults (ITSC and USV) where using DWER indicators as NN input. The results obtained of combination between the DWER and NN are effective and proved its ability to detect both faults under different load conditions and distinguish between them accurately with low error (10^-5).