WAVELET ANALYSIS OF PLANETARY GEARBOX VIBRATION DATA FOR FAULT DETECTION AND DIAGNOSTICS

摘要

Vibration data analysis plays a critical role in condition monitoring for fault detection, diagnostics and condition based maintenance (CBM) purposes. This requires development of effective methods for analyzing vibration data. A literature research has shown that while there exists a body of literature on vibration data analysis techniques for gearboxes with fixed-axis gear sets, techniques for analyzing vibration data for gearboxes with planetary gear sets have been less explored. The aim of our research has been to develop such a method for analyzing vibration data of planetary gears. In this paper, an approach based on the application of wavelet transform to the time synchronously averaged (TSA) signal is presented. We have compared FFT and some fault indicators of the raw signal, TSA signal, continuous wavelet transform (CWT) and discrete wavelet transform (DWT) applied to the TSA signal. It should be noted that the analysis of planetary gearbox vibration data is significantly more complex than the analysis of fix-axis gear sets data. We have found that the methods using wavelet transform can achieve desirable feature extraction for fault detection, diagnostics and CBM purposes.