A Novel application of Wavelet-based Bispectrum Analysis to Diagnose Faults in Gears

摘要

Aimed at the non-stationary and non-Gaussian nature of the signals generated during the condition monitoring and diagnosis of the performance of gears, a novel proactive strategy for detecting the incipient faults developed in gears is put forward in this paper. This is accomplished in two ways. Firstly, the original signal is decomposed using wavelet transformation followed by the application of bispectrum analysis. This methodology not only integrates the merits of both the analytical techniques, but also it offers an efficient tool to extract the fault features of the gear under investigation. The findings obtained from the results of this investigation are promising.