Empirical mode decomposition (EMD) and Hilbert spectral analysis (HSA) representa desperate attempt to break the suffocating hold on data analysis by the twinassumptions of linearity and stationarity. To analyze the data from nonlinear andnon-stationary processes, various attempts such as Spectrograms, Wavelet analysis,and the Wigner-Ville distribution have been made, but the EMD-HSA approach isunique and different from the existing methods of data analysis. The EMD-HAS istruly an adaptive time-frequency analysis. It does not require an a priori functionalbasis. Instead, the basis functions are derived adaptively from the data by the EMDsifting procedures; the instantaneous frequencies are computed from derivatives ofthe phase functions of the Hilbert transform of the basis functions; the final resultis presented in the time-frequency space. The EMD-HSA is a magnifying glass foranalyzing the data from nonlinear and non-stationary processes. The EMD-HSAresults are intriguing and are no longer shackled by spurious harmonics (the artifactsof imposing a linearity property on a nonlinear system) or limited by theuncertainty principle (the consequence of Fourier transform pairs in data analysis).EMD-HSA was originally designed in 1995 specifically to study water surfacewave evolution, the phenomenon of high frequency waves with short fetch evolvinginto low frequency waves at long fetch. With the EMD-HSA method, it wasfound that the evolution of the waves was not continuous but abrupt, discrete andlocal. Subsequently, NEH spent two years visiting Caltech at the invitation of ProfessorTheodore Y. Wu. Under the guidance of Professor Wu and Professor OwenM. Phillips of the Johns Hopkins University, the EMD-HSA method was furtherdeveloped and various applications explored. Professor Wu designated the methodas the Hilbert-Huang Transform (HHT), a name later adopted by NASA to avoidthe awkward name of EMD-HSA. It is only fair to say that the HHT would nothave been developed without the encouragement and guidance of Professors Wuand Phillips
Empirical mode decomposition (EMD) and Hilbert spectral analysis (HSA) representa desperate attempt to break the suffocating hold on data analysis by the twinassumptions of linearity and stationarity. To analyze the data from nonlinear andnon-stationary processes, various attempts such as Spectrograms, Wavelet analysis,and the Wigner-Ville distribution have been made, but the EMD-HSA approach isunique and different from the existing methods of data analysis. The EMD-HAS istruly an adaptive time-frequency analysis. It does not require an a priori functionalbasis. Instead, the basis functions are derived adaptively from the data by the EMDsifting procedures; the instantaneous frequencies are computed from derivatives ofthe phase functions of the Hilbert transform of the basis functions; the final resultis presented in the time-frequency space. The EMD-HSA is a magnifying glass foranalyzing the data from nonlinear and non-stationary processes. The EMD-HSAresults are intriguing and are no longer shackled by spurious harmonics (the artifactsof imposing a linearity property on a nonlinear system) or limited by theuncertainty principle (the consequence of Fourier transform pairs in data analysis).EMD-HSA was originally designed in 1995 specifically to study water surfacewave evolution, the phenomenon of high frequency waves with short fetch evolvinginto low frequency waves at long fetch. With the EMD-HSA method, it wasfound that the evolution of the waves was not continuous but abrupt, discrete andlocal. Subsequently, NEH spent two years visiting Caltech at the invitation of ProfessorTheodore Y. Wu. Under the guidance of Professor Wu and Professor OwenM. Phillips of the Johns Hopkins University, the EMD-HSA method was furtherdeveloped and various applications explored. Professor Wu designated the methodas the Hilbert-Huang Transform (HHT), a name later adopted by NASA to avoidthe awkward name of EMD-HSA. It is only fair to say that the HHT would nothave been developed without the encouragement and guidance of Professors Wuand Phillips