Condition Monitoring of Gears and Advanced Signal Processing Techniques towards More Effective Diagnostic Schemes

Abstract

A diagnostic methodology of artificial defects in a single stage gearbox operating under various load levels and different defect states is proposed in the present work based on vibration recordings as well as advanced signal analysis techniques. Two different wavelet-based signal processing methodologies, using the discrete as well as the continuous wavelet transform, were utilised for the analysis of the recorded vibration signals and useful diagnostic information were extracted out of them. Both wavelet analysis techniques provided the ability of distinguishing between the healthy and the artificially defected gears. In this way, the health monitoring potential of vibration monitoring in the case of rotating machinery and gearboxes obtains a new dynamic under the prism of sophisticated time-frequency signal processing schemes, rather than conventional FFT-based approaches.

Application of psychoacoustics for gear fault diagnosis using artificial neural network

“…When in operation, the major causes of failure of the gearboxes are cracked gear tooth, pitting, wear, etc. Wang et al, 20 Saravanan et al, 21 El Badaoui et al, 22 Wang et al 23 and Loutas et al 24 have demonstrated use of intelligent techniques for effective classification and gear fault diagnosis.…”

De-Noising Algorithm for Magnetotelluric Signal Based on Mathematical Morphology Filtering

“…If time for observation is long enough, the effect of lowfrequency part improvement is also obvious, while the humanities interference can be less inhibited [2]. Wavelet de-nosing method is completely dependent on selection of wavelet functions, sometimes as the scale increases, the corresponding orthogonal basis functions localized variation of the spectrum [3,4], making the more sophisticated decomposition of MT signals is limited. Although this method can improve a single type of noise and strong energy data, when noise energy is weak and number is large, the capacity of noise identification and de-noising is often very poor [5].…”