A Study About The Application And Types Of Mother Wavelet Function Applying On The Fault Diagnosis And Detection Of Rotating Machinery

In this thesis, I mainly introduced the following few problems about the basic principle of wavelet analysis and its application:Firstly, briefly introduced the important status of rotating machinery, major methods of vibrating diagnosis and its present development; primarily described basic principle and prominent advantages of one of the time-frequency analysis — wavelet analysis, and compared the characteristics and features between popular different wavelet functions. Simulated a large amount of major and typical fault signal of rotating machinery, applied wavelet analysis and traditional FFT on these signals and compared them. Fully revealed the advantages and promising prospects of wavelet analysis.Secondly, de-noising is one of the main application field of wavelet, this thesis analyzed several typical noised fault signals by means of MATLAB, a few conclusion was achieved about the influencing factors of wavelet de-noising, such as mother wavelet(compact supporting, length of filter, symmetry, moment of vanishing ), noise type, threshold, wavelet de-composition level, WT and WPT. At the end an adaptive wavelet de-noise method was proposed, and obtain a relative nice effect, which will promote its application.Lastly, it introduced basic principle, type, feature of artificial neural network, through the analysis of several application cases of the combination of wave and ANN. it discussed major way and key element of realization of this technology combination, finally it gave some opinions and prospect of my own.