A comparative study between the MCSA, DWT and the vibration analysis methods to diagnose the dynamic eccentricity fault in induction motors

This paper presents three interesting methods that are widely used in the diagnosis of electrical machines. Today, many researchers use a technique to detect the faults which may cause the induction machines. The choice of one method over the other is essentially depends on its effectiveness, ease of use and speed. Early detection of faults is a valuable advantage. Our new study is to investigate the efficiency of each method under a dynamic eccentricity fault (DE) in order to detect the characteristic signatures. Motor current signature analysis (MCSA) method is widely used in the diagnosis; it based on analyzing the fast Fourier transform (FFT) of the stator current. Another method which is opens the door to analyze signals in simultaneous time-frequency. It is called the discrete wavelet transform (DWT) which has the ability to allow for studying transient phenomena and non-stationary signals. We will calculate and analyze a particular frequency band of the stator current by DWT coefficients. Our study also treats the details and approximates by Daubechies order. We will present an important indicator of this technique; it's the energy of some details. The last technique is the vibration analysis that is widely used now in industry to detect the faults in electrical machines. The vibration is the result of several causes such as mechanical or electromagnetic. The general objective of the vibration analysis is to analyze the signals to study the content. In this study, we will compare between three methods currently used by exploiting the experimental results to analyze the induction machines faults. The efficiency of each technique was discussed in order to reach a wise decision.