On the selection of informative wavelets for machinery diagnosis

Abstract

A new method of machinery fault diagnosis based on wavelet analysis is presented. We introduce an extension to Mallat and Zhang's matching pursuit for machinery diagnosis is presented. Instead of the ‘best matching’ criterion, a mutual information measure is used to search a redundant wavelet dictionary for a small set of wavelets that carry meaningful information about machinery faults. With these informative wavelets treated as feature extractors, this approach effectively facilitates the diagnosis of machinery faults of a non-stationary nature. This method has been applied to the detection of diesel engine malfunctions. The results show that both the sensitivity and the reliability of this approach are good.