Time-frequency strain data analysis of suspension using the Hilbert-Huang transform / Nadia Nurnajihah Mohamad Nasir ... [et al.]

This paper aims to study the application of the Hilbert-Huang transform in automotive component strain data. The objective is to analyse time-frequency strain data and investigate specific and indicative behaviour patterns of the time-frequency parameters by using the Hilbert-Huang transform. Hilbert-Huang transform is different from the traditional Fourier transform, which is used only for linear and stationary signals analysis. Fourier transform is different if compared with the Hilbert-Huang transform. Hilbert-Huang transform is designed to analysing the nonlinear and non-stationary signals and a more suitable tool for this kind of system. Empirical mode decomposition can characterise the intrinsic mode function to decompose the signal by mean of the time-frequency variations of signals. The empirical mode decomposition extracts both the original signals into a set of intrinsic mode functions which emphasises different oscillation mode with different amplitudes and frequencies. The intrinsic mode functions component produces significant and more effective physical analysis in the physical process at different time scales. The results obtained can also be observed from numerical parameters that there are difference between the wide inter-subject differences in the variance and the contribution period of each signal mode in intrinsic time-frequency to the total number of signal content. The mean period for both first decomposition signals is ~2 and ~6. Reconstruction of new signal is done using the result of decomposition signal, intrinsic mode functions and the residue. The reconstruction signals have a difference in the maximum amplitude less than 1.136×10-13 and 2.273×10-13 that indicate unknown noise. This study represents the decomposition signal which was at high frequency in the histogram of Kernel estimation probability based on the strain data signal in the automotive component.