The Discussion Of Adaptive Signal Decomposition Based On Instantaneous Frequency And Local Narrow Band Signals

Signal analysis and processing has been one of the most active areas of research, In recent years, the development of non-stationary signal analysis, especially concerns a lot of people. There have been more new theory and methods around signal analysis and processing which have been applied into a wider range of fields. It is no exaggeration to say that non-stationary signal processing technology has greatly promoted the development of signal processing disciplines.First, this paper introduces some basic theories and related concepts of signal analysis and processing, and briefly introduces some of the traditional processing methods of non-stationary signals, such as Short-Time Fourier Transform,Wavelet Analysis,Gabor Transform. These methods are dependent on the selection of basis functions, which leads to the studies and researches of the adaptive signal decomposition.Furthermore, introduces the empirical mode decomposition (EMD) algorithm, which proposed by N.E.Huang. After ten years of development, its theory and technology is relatively mature, and the application is also more extensive. This paper mainly introduces the basic principles and the decomposition process of EMD method, as well as its problems and shortcomings.The adaptive signal decomposition algorithm based on local narrow band signals proposed by S.L.Peng, is a new method of non-stationary signals after the EMD method, which has some advantage beyond the EMD method in some respects. This algorithm can guarantee the convergence, and the parameters involved in the adaptive algorithm can be obtained by itself, instead of subjective experience. By comparing with the Matching Pursuit (MP) method, it also gives their connection.Based on S.L.Peng's first type of local narrow band signal decomposition, this paper proposes an optimization model, which can extract the local narrow band signals quickly and accurately. It explains the mechanism of this optimization model, and constructs a fast algorithm to solve the model. Finally, through specific numerical tests, it shows the feasibility and effectiveness of the optimization model.