Fatigue road signal denoising process using the 4th order of daubechies wavelet transforms

Abstract

This study explores an application of the wavelet denoising technique in a fatigue road load variable amplitude data set. In this study, the wavelet denoising application has been implemented using the 4th order of Daubechies family, with the adaptation of fifteen levels decomposition process. From the view of current research trend, the wavelet-based denoising approach is widely used using vibration random signal, but it is rarely been used in the scope of fatigue road loadings, or also known as fatigue strain signals. The idea of this study came from the some previous vibration analysis research and it was found to be suited to the approach of fatigue signal denoising process. High amplitude events in a fatigue road signal are very important and they should be retained because of these features caused significant damage of the components, particularly in automotive applications. After the fatigue signal has been denoised, the global signal statistical calculation and fatigue damagel1ife analysis were performed in order to validate the applicability of this denoising technique. From the analysis, it was found that the wavelet denoising approach was not suitable to analyse fatigue data and the major concern is the omission of high amplitude events from the original road loading, hence to a significant fatigue damage difference when compared to the edited road loading. © 2008 Asian Network for Scientific Information.