Testing the Effectiveness of Wavelet-based Denoising Schemes for Gear Fault Diagnosis

Main Article Content

Mustapha MERZOUG

Abstract

Vibration analysis has always been used for the diagnosis of gear faults. The vibration signals obtained are generally contaminated with noise and therefore not usable. The denoising of the vibration signals obtained helps in the diagnosis of faults, giving significant results. The wavelet transform increases the signal-to-noise ratio, reduces the mean absolute error, and is recommended for denoising gear vibration signals. The extracted signals must be denoised by choosing an appropriate denoising scheme to preserve the information contained in the signal. An approach has been developed in this work to show the effectiveness of this method. We chose a single-stage dynamic model with basically two shafts, one driving a load and the other a motor, to obtain this signal. Four bearings and two gears support both shafts. The meshing stiffness and the response have been calculated by an analytical method and a Newmark integration scheme, respectively. Then, we deliberately introduced a defect in the shape of a crack in a gear tooth. We selected Daubchies wavelets, which are well adapted to this type of problem. The objective is to control the different parameters related to our approach based on the wavelet transform, which is the levels of decomposition and the order of the wavelets. This approach consists of observing a series of indicators for a series of wavelet orders depending on the severity of the defect. The novelty of this research work is the improvement of scalar indicators while mastering the parameters linked to the wavelet transform. Our approach has been compared to other well-known methods in the literature.

Downloads

Download data is not yet available.

Article Details

How to Cite
[1]
2023. Testing the Effectiveness of Wavelet-based Denoising Schemes for Gear Fault Diagnosis. Romanian Journal of Acoustics and Vibration. 20, 2 (Dec. 2023), 122–129. DOI:https://doi.org/10.1234/xde0wv62.
Section
Articles

How to Cite

[1]
2023. Testing the Effectiveness of Wavelet-based Denoising Schemes for Gear Fault Diagnosis. Romanian Journal of Acoustics and Vibration. 20, 2 (Dec. 2023), 122–129. DOI:https://doi.org/10.1234/xde0wv62.

Most read articles by the same author(s)

1 2 3 4 5 6 7 8 9 10 > >>