Abstract
This paper present the use of wavelet transforms for denoising a road load fatigue strain data set. For the purpose of this paper, the 4(th) order of Daubechies wavelet with fifteen decomposition levels was adopted. The denoised approach of a random data was seldom used in the noise and vibration research, but it rarely been used in the scope of fatigue data analysis. This study adopted the approach used in the previous related research, which was the application in a vibration analysis. It was hypothesized that this denoising method may be suited for the analysis of fatigue road load data, seems this data type is having similar pattern to the vibration random data. For the fatigue life assessment, high amplitude cycles in the fatigue data set are vital to be retained after the denoising process, as these cycles gave higher fatigue damage potential. After the fatigue signal being denoised, the statistical and fatigue analysis were performed in order to validate the applicable of this technique is applicable in fatigue signal analysis. From the analysis of this paper, it was found that the wavelet denoising approach, by means of the 4(th) order of Daubechies wavelet, was not suitable to analyse fatigue random data because. It is because of this approach diminished high amplitude cycles from the original signal.