Time-frequency Analysis Of Theoretical And Applied Research. Of Emd

With the development of science and technology, nonlinear,nonstationary signal processing and its application in engineering gain more and more attention. As we all know, nonlinear,nonstationary signal processing is a very difficult problem not only in signal processing field but also in fault diagnosis. Many technologies have appeared to meet the need of processing nonlinear,nonstationary signal. Empirical mode decomposition (EMD) method is one of them. After several years' research and application it is approved that EMD is a wonderful way in nonlinear,nonstationary signal processing. But EMD is a new technology, and much research has to be done for testifying its validity. Base on the project I took part in during my three years'study, EMD method is deeply studied both in theory and application in this paper. The main contents are illustrated as follows:(1) The basic concepts of instantaneous frequency and intrinsic mode function are illustrated, research has been done on the algorithm of EMD and the discussion of the feature of it is followed. Secondly, the capabilities of Hilbert spectrum are studied in this paper. Compared to short time Fourier transform, Wigner-Ville distribution and wavelet spectrum, theoretic analysis and simulation experiment results show that EMD spectrum has perfect time-frequency concentration, and good time-frequency resolution characteristics. Besides, the current hot topics of it are discussed.(2) Based on the time-scale filtering property of EMD, the filter banks are designed and applied to extract the trend of a signal. Time-frequency analytic method based on EMD is used to analyze time-frequency characteristics of numerical results of classical modulation signal and nonlinear equation system. It is used for Boeing 737 turbine vibration fault diagnosis. The low frequency component is detected as well as other useful turbine vibration signal components.(3) Bidimension EMD (BEMD) is investigated in this paper. This method can extract components with different frequencies. The simulation results prove that the detail information of the image can be obtained by BEMD method. Lastly, BEMD is applied to digital watermarking technique and the whole watermarking scheme based on BEMD is proposed. The watermark is inserted and extracted successfully and the robustness of this scheme is tested. This new watermarking scheme based on BEMD offers a new way of digital watermarking technology.