Based On Wavelet Transfer Rotating Mechanical Fault Diagnosis

The difference of base function cause the difference of observation angle and observation method for signal, this is the essence of Wavelet transfer differentiate Fourier transfer f Short-time Fourier transfer, comparing performance and characteristic of various wavelet base functions in common use in fault diagnosis and studying inner connection of different fault diagnosis signal and various wavelet base functions, it is pointed out that genuine optimum wavelet transfer is that signal decompose of base function should be most close to and be most suited to signal feature .It is indicated that Hermitian wavelet has the ability to sensitively recognize the transient phase of signal through comparing the advantage and shortcoming of many kinds of methods used for singularity detection of signal. The time-scale amplitude plot and phase plot based on Hermitian wavelet transform are presented to singularity detection of signal. The corresponding relation relation between singularity and amplitude plot or phase plot is created. As successful application, fault of axletree damnify is found by means of this method.