Research On Fault Diagnosis Technologies Of Asynchronous Motor Based On Transient Signal Analysis

Asynchronous motor, with its high reliability and the advantages of strong ability to adapt, is widely used in industrial production and life as the main power unit. Asynchronous motor running state is directly related to the overall system performance and security. Therefore,running state identification and fault diagnosis of asynchronous motor has been one of the key fields which scientific research workers and engineering and technical personnel pay attention to.As the current signal of motor running state is one of the important means of fault diagnosis analysis, the current signal contains abundant motor working status information,and using the current signal fault diagnosis of motor is a kind of effective method. But with the electrical, mechanical and hydraulic system facing development trend of more higher, more precise, and more advanced, system gradually reflects the strong coupling and nonlinear characteristics, and often be with load of extreme condition of continuous change. In this case, the traditional MCSA spectrum is characterized by a fuzzy effect, and it is difficult to extract the fault features. Therefore, the article obtains characteristics from the motor start signal by using Fourier Bessel algorithm for motor transient signal, and the typical motor fault feature is extracted. And classify extracted feature by the principal component analysis.Finally forming based on transient current signal analysis method of asynchronous motor fault diagnosis. Full text main content is as follows:(1) On the basis of asynchronous motor structure and working principle, with multi loop theory, establish a theoretical model of normal condition asynchronous motor; On the basis, analysis fault mechanism of the stator turn-to-turn short circuit and rotor broken bar, and establish its theoretical model.(2) Based on the theoretical model, research load robustness of typical conditions of motor steady state signals and transient.(3) Using Fourier Bessel algorithm, analysis motor starting current, to extract the fault signal characteristics, and comparing with the traditional wavelet analysis.(4) Using PCA algorithm to classify and recognize the fault.