The Research Of Diagnosis In Motor Fault Based On HHT Transformation

With the development of modern science and technology, and the improvement of electric degree, asynchronous motors play a more and more important role in modern industrial plants. When it work out of order or off-the line, not only the motor itself is destroyed but also the whole system would be in danger and the enormous economic losses would be caused. Through monitor and analyze the state of motor, the fault can be found earlier, further deterioration of failure be prevented and accidents caused by unexpected production losses be reduced, as well as to create conditions to achieve a state of repair.As the fault signals of electric machine are non-stationary transient ones, the traditional signal analysis methods, such as Fourier Transform,Gabor Transform,Wavelet Transform,Wigner-Ville/Cohen distribution etc, are not so efficient and useful for the fault signal extraction. They are the improvement methods which base on the Fourier analysis by introducing localized function to deal with the dilemma of global integral transform. In essential they are some kind of localized Fourier analyses so as to suffer from the restrict requirements of the linear and the stationary process, unfortunately, non-stationary and nonlinear signals are not uncommon in the real world. Hilbert-Huang transform (HHT) gets rid of the background of Fourier analysis and is taken as the tremendous breakthrough of time-frequency analysis in recent years.Firstly, the research status,notion,method of diagnosis in asynchronous motor fault,the principle and common faults of asynchronous were introduced. Secondly, Hilbert-Huang Transform and its two important concepts were presented, the method of extension of extreme value was used to empirical mode decomposition (EMD) of the boundary points, which reduced the influence of boundary effect. And then, the common faults'features and the formation mechanism of squirrel cage induction motor (SCIM) was analyzed. Finally, a new method using HHT for the analysis of faulted motor signal by the principle of asynchronous motor. This method made HHT used to diagnosis in squirrel cage induction motor (SCIM) fault, the feature frequency of fault signal was obtained through the analysis of dynamic and static fault signal of motor. The results indicated that the three mean instantaneous power approach can clearly separate feature frequency by analyzing the three mean instantaneous power and signal-phase current ,a new method named frequency-gravity way was given which used to further process partial Hilbert spectrum, furthermore, the signal spectrum which would be more consistent with realities was acquired. The results of the experiment showed that this method got better effect.