Applications of empirical mode decomposition for processing nonstationary signals

Empirical mode decomposition (EMD) has recently been pioneered by Huang as a fully data-driven technique aimed at decomposing nonstationary signals in a set of “Intrinsic mode functions” (IMFs, Empirical modes). We will report on the main theoretical aspects of EMD, its extensive possibilities, and various contemporary applications. We will pay attention to detrending; denoising; Hilbert-Huang time-frequency analysis; and a very perspective and actual scientific direction known as Data Mining, which involves such problems as segmentation, cluster-analysis (clustering), classification, etc. The text was submitted by the authors in English. Klionski Dmitrii Mikhailovich. Born 1986. Bachelor of Engineering and Technology in Telecommunications. Programming engineer. Work in the field of applied mathematics, programming 3 years. Member of Regional St. Petersburg Division of RAPRIA. Scientific interests: adaptive signal analysis, methods of digital spectral analysis, statistical signal processing, Matlab. Author of 15 publications in the field of image recognition and analysis. Geppener Vladimir Vladimirovich. Born 1949. Received Doctoral degree (in Technical Sciences) in 1996. Leading researcher, Public Corporation "Scientific Engineering Center of St. Petersburg Electrical Engineering University", Professor of St. Petersburg State Electrical Engineering University. Member of Regional St. Petersburg Division of RAPRIA. Work in the fields of applied mathematics, programming over 30 years. Scientific interests: wavelet analysis, speech recognition, hydroacoustics, methods of digital spectral analysis, Matlab. Author of more than 50 publications. Oreshko Nikolai Ivanovich. Received Candidate’s degree (in Technical Sciences) in 1980. Leading researcher, Public Corporation “Scientific Engineering Center of St. Petersburg Electrical Engineering University”. Work in the fields of applied mathematics, programming over 30 years. Scientific interests: wavelet analysis, statistics, methods of digital spectral analysis, Matlab. Author of more than 50 publications. Vasiljev Aleksandr Vladimirovich. Born 1950. Received Candidate’s degree (in Technical Sciences) in 1996. Head of Department, Public Corporation “Scientific Engineering Center of St. Petersburg Electrical Engineering University”. Work in the fields of applied mathematics, programming over 30 years. Scientific interests: wavelet analysis, speech recognition, hydroacoustics, methods of digital spectral analysis, Matlab. Author of more than 50 publications.