On performance differences of EMD and WD in the nonlinear time series analysis

The procedure of the Hilbert transformation (HT) based on empirical mode decomposition (EMD) is first decomposing a nonlinear time series into its intrinsic mode functions (IMFs), and then taking the HT of each IMF and computing the instantaneous amplitude and frequency. In the context of an ideal time series, merits and defects of EMD and wavelet decomposition (WD) as well as HT and wavelet transformation (WT) in the nonlinear time series analysis are systematically analyzed/compared in this paper, and aiming at their defects, some proposals for possible improvement are also given. Research results show that the combination with the EMD-based analysis method and the WD-based one may more effectively identify the characteristic information of the original time series.