A Study Of Ensemble Empirical Mode Decomposition

A time-frequency analysis method, Hilbert-Huang Transform (HHT), which was proposed by Huang et al, is more and more popular in recent years. Based on the local characterizations of the signal, this method is especially developed for adaptively analyzing non-stationary and non-liner data. The key part of the method is the empirical mode decomposition, EMD, with which any complicated data can be shifted into a small number of intrinsic mode functions . Though this method has already been applied into many fields, there are still some problems. In this paper, the mode mixing problem has been discussed. Then the ensemble empirical mode decomposition method, EEMD, which was considered as an improvement of the original empirical mode decomposition method, has been introduced.In the application study of the EEMD, a speech signal has been analyzed using EEMD to demonstrate that the ensemble empirical mode decomposition method do a good work of eliminating mode mixing. Moreover the data from the experiment of hydraulic-turbine cavitation has also been studied using EEMD and Hilbert marginal spectrum. Compared to the Fourier spectrum, a new feature of the relationship between the cavitation and noise density was obtained through EEMD and Hilbert marginal spectrum.