FCM-based Optimisation to Enhance the Morlet Wavelet Ability for Compressing Suspension Strain Data

The study aims to enhance the ability of the wavelet-based extraction for fatigue life assessment. A SAE-owned fatigue strain random signal, called SAESUS was extracted using the Morlet wavelet and produced non-damaging and damaging segments. Furthermore, the segments were clustered using the Fuzzy C-Means method in order to analyse the segment behaviours. The clustering method scattered the segments based on the difference in the root-means square, kurtosis and fatigue damage values. Damaging segments then were assembled together in order to have a new edited signal. The extraction process was able to shorten the original signal up to 41% and it was able to retain at least 90% of both statistical parameters and the fatigue damage. Finally, it is suggested that the Morlet wavelet successfully identified the higher amplitudes in the strain data.