作者: Wenyi Liu , Baoping Tang
DOI: 10.1016/J.ESWA.2010.12.107
关键词: Morlet wavelet 、 Mathematics 、 Wavelet packet decomposition 、 Pattern recognition 、 Noise (signal processing) 、 Artificial intelligence 、 Filter (signal processing) 、 Continuous wavelet 、 Window function 、 Wavelet 、 Second-generation wavelet transform
摘要: Research highlights? The hybrid time-frequency method is based on Morlet wavelet and auto terms window. ? parameters are optimized by cross validation minimum Shannon entropy method. useless noise in the original signal can be filtered CWT filter de-noising process. Wigner-Ville Distribution suppressed ATW. In this paper, a (HTM) improved window (ATW) presented. wavelet, for its shape similar to mechanical shock signals, added two which decide of mother wavelet. appropriate scale parameter continuous transformation (CWT) calculated using (CVM) An ATW Smoothed Pseudo (SPWVD) spectrum designed as function suppress (WVD). gear fault diagnosis experiment results show that proposed has good de-nosing performance effective removing extracting feature.