Research On The Detection Methods Of Rotor And Bearing Fault In Indution Motors

Induction motors can be used in almost all kinds of complicated work environments, and become the important components of many industrial processes. They have been widely used in many industrial fields. It is a kind of coverage broadest and usage of the biggest motor. Induction motor can guarantee the normal operation of the process of production efficiency, safety, promptness. Motor failure will not only lead to the destruction of production facilities, affecting the conduct of production, or even stop production, but also endanger the safety of workers, resulting in huge economic losses. Therefore, fault detection of induction motors especially at an early stage is necessary for ensuring the normal operation of the industry processes.This paper proposes a new fault detection method based on spectrum analysis of Hilbert modulus, which combines Wavelet Transform and Hilbert transform to detect broken-rotor-bar faults of induction motors. Hilbert transform is introduced into Wavelet analysis to obtain analytical results of wavelet transform of the signals. Based on analytical results of wavelet transform, this method calculates the location of fault signature and eliminates the effect of the oscillation. The Hilbert modulus is defined as the sum of the square of a signal and its conjugation. It is used to covert the fundamental component in original phase current to the direct current component and to convert the fault characteristics frequency of broken-rotor-bar,(1-2s)f1 to current component of 2sf1 frequency. This method overcomes the shortcomings of the fault signatures often being concealed by the fundamental frequencies. The results of simulation demonstrate that the proposed method can overcome the drawback of the traditional current spectrum analysis method, and provides an accurate fault diagnosis.The other research done in this paper is focused on combining empirical mode decomposition (EMD) and wavelet analysis. EMD performs well for self-adaption decomposing signals and it can be used for extracting features of the signals. Therefore, EMD and wavelet analysis are combined to analyze current signals from bearing fault motors in this paper. The efficency of the method is demonstrated by the simulation results.Finally, this paper sums up the proposed methods and shows the development prospects of fault detection methods for induction motors.