A Study on Spiral Bevel Gear Fault Detection Using Artificial Neural Networks and Wavelet Transform

Abstract:

Article Preview

Based on normal and defective gears of spiral bevel gear pair test, a study is represented to develop the performance of gear fault detection with artificial neural networks and wavelet transform. In order to research the relevant studies of gear failures, a gear fault test rig is designed and constructed, with which vibration test are processed for collecting the signals of a gearbox from this rig. The noise is removed from the original time-domain vibration signals by application of wavelet analysis threshold technique. The extracted energy features from those preprocessed signals are implemented by the wavelet transform, which are used as inputs to the artificial neural networks for two-pattern (normal or fault) recognition. The results show that the represented recognition accuracy of the ANN and WT method for gear fault diagnosis is 100% that is much higher compared with the results of application of ANN separately.

Info:

Periodical:

Edited by:

Zeyong Yin, Chengyu Jiang, Datong Qin, Peixin Qiao and Geng Liu

Pages:

214-217

DOI:

10.4028/www.scientific.net/AMM.86.214

Citation:

B. B. Fu and Z. D. Fang, "A Study on Spiral Bevel Gear Fault Detection Using Artificial Neural Networks and Wavelet Transform", Applied Mechanics and Materials, Vol. 86, pp. 214-217, 2011

Online since:

August 2011

Export:

Price:

$35.00

In order to see related information, you need to Login.

In order to see related information, you need to Login.