Saad Al-Baddai and Karema Al-Subari and Elmar Lang and Bernd Ludwig
Optimizing Approach for Sifting Process to Solve a Common Type of Empirical Mode Decomposition Mode Mixing
717 - 725
2017
11
6
International Journal of Electrical and Computer Engineering
https://publications.waset.org/pdf/10007433
https://publications.waset.org/vol/126
World Academy of Science, Engineering and Technology
Empirical mode decomposition (EMD), a new
datadriven of timeseries decomposition, has the advantage of
supposing that a time series is nonlinear or nonstationary, as
is implicitly achieved in Fourier decomposition. However, the
EMD suffers of mode mixing problem in some cases. The aim of
this paper is to present a solution for a common type of signals
causing of EMD mode mixing problem, in case a signal suffers
of an intermittency. By an artificial example, the solution shows
superior performance in terms of cope EMD mode mixing problem
comparing with the conventional EMD and Ensemble Empirical
Mode decomposition (EEMD). Furthermore, the oversifting problem
is also completely avoided; and computation load is reduced roughly
six times compared with EEMD, an ensemble number of 50.
Open Science Index 126, 2017