2.5. Signal Processing in the Time-Frequency Domain

Since extensively used FT is restricted to stationary signals, it is unsuitable to process the fault signals of machinery which are nonstationary when a fault occurs. Moreover, the frequency spectrum acquired by FT cannot reflect the change of each frequency with time. To overcome these disadvantages, time–frequency analysis techniques are developed to analyze a nonstationary signal from time–frequency distribution. Up to now, classical time–frequency analysis techniques mainly include STFT, WVD, wavelet transform (WT), even wavelet packet transform (WPT), and HHT, etc. These techniques can reflect the change of a signal from both time domain and frequency one. Moreover, they can disclose intrinsic ...

Get Intelligent Fault Diagnosis and Remaining Useful Life Prediction of Rotating Machinery now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.