Subband-based EMD Signal Decomposition Algorithm

In modern signal processing, the development of the non-linearity and non -stationary signal processing methods are especially attractive. The appearance of non -linearity and non-stationary signal processing method makes people process this kind of signals not depend on the FFT transform only. Furthermore, the result of processing is more verity and significance. Hilbert-Huang Transform (HHT) is a well-known time -frequency analysis technique. The key part of the method is the'empirical mode decomposition (EMD)', and based upon which any complicated data set can be decomposed into a finite and often small number of'intrinsic mode functions'that admit well-behaved Hilbert transforms. This decomposition method is adaptive and highly efficient. The main content of the thesis is as follows:Firstly, based upon the principle of the non-stationary signals processing, the EMD is illustrated in detail. In addition, the physical properties of Hilbert spectrum and marginal spectrum are discussed. Moreover, the hot topics on the EMD are mentioned.Secondly, the two algorithms of the complex signal EMD are reviewed. The principle of the bivariate EMD is summarized and its efficiency is validated by simulated experiments.Finally, the M-band multirate filter banks are reviewed. Considering that the bivariate EMD brings some false components and the distortion of signals in stronger noise background, we combine the M-band filter bank and the bivariate EMD algorithm to construct a new subband-based EMD algorithm. Simulated experiments show that the subband-based EMD algorithm is an efficient way to reduce the effect of noise and it achieves better performance than the bivariate EMD algorithm.