| There is a new method for processing non-linear and non-stationary signal,Hilbert-Huang Transformations(HHT). This method can give the time-frequency distribute of non-stationary signal. Therefore, as soon as put forward, this method become numerous quickly got an extensive application at each realm.In this paper, the HHT method was applied to analyses ordinary signals. The result is that EMD method is self-adaptive, dose not need advanced bases. But the influence of the noise is very greatly, decline the precision of decomposition of the EMD very much, so that can not get clear Hilbert spectrum. The precision of decomposition can be improved by declining noise by wavelet. And this method also can get clear Hilbert spectrum. Through the Wuhan University of science and technology′s gear fault laboratory bench applying, the author gathers several kinds of gear fault signals, and carries on an analysis of gear vibration signal in four modes: normal, wearing, broken and circular pitch error. Based on noise reduction by wavelet and HHT, we can abstract and find out the various wheel gears'fault character by Hilbert spectrum. Power spectrum analysis confirms that the Hilbert spectrum is very accurate. On this foundation, the Hilbert spectrums of various typical gears'fault are the standard spectrum so that can be used to diagnosing the gear fault.In one side, it is a new and powerful tool that applying the Hilbert-Huang Transformation method to diagnose the wheel gears'malfunction. It is also a new time-frequency analysis method. This method can give clear time-frequency spectrum but need very shot time-sequence, which can accurately diagnose the gears'malfunction. In the other side, Empirical Mode Decomposition is self adaptive, doesn't need advance base, because its decomposition is based on the signal's local time scales which is unexampled. These advantages come to a decision rare that transformation method to have extensive applied foreground in the mechanical fault diagnosis.
|
PDF Full Text Request