Fast Filtering Decomposing Signal Into Intrinsic Mode Function And Its Applications In Marine Data Analysis

The vastness and changefulness of ocean render the acquisition of oceanic signals hard and difficult. It is thus of crucial importance for marine research to reasonably analyze oceanic signals acquired, so as to extract useful information as much as possible from them.In this thesis, after brief reviews on the Fourier analysis, short-time Fourier analysis and wavelet analysis, and a more detailed review on the EMD (empirical mode decomposition) analysis recently presented by Huang et al.(1998), a new method for the EMD is presented. To show its validity, the new method is used for analyzing some typical oceanic signals and the results are reported.In contrast with the EMD method presented by Huang et al.(1998), the new method has the following advantages:1) With a certain restriction, the intrinsic mode function ( IMF ) set obtained from a signal by the new method is unique, and the number of IMFs in the set is minimum;2) The IMFs obtained by the method are strictly orthogonal each other;3) The method is fast and high efficient. Because of these advantages, the new method is not only more strict in theory but also more competent and convenient in practical use.