The Spectral Decomposition Technique Based On MPD

The spectral decomposition is the basis of seismic data processing. Seismic signal is complex because of non-stationary signals, the appropriate approach is to ensure that time-frequency analysis spectral analysis of seismic data when the key to the quality.In this paper, more comprehensive study of the basic principles of time-frequency analysis and development, and has compared with several basic time-frequency analysis methods, such as short time Fourier transform, S transform, continuous wavelet transform and Wigner-Ville distribution.Then discusses in detail the spectral decomposition of the ideological sources of MPD (Matching Pursuit Decompositon), the basic principles and operations to achieve. The MP algorithm has good time-frequency clustering, consider that MPD algorithm based on spectral decomposition technology can have a better performance in seismic signal time-frequency distribution information, and can describe the formation response to different frequencies. It could provide an effective means of hydrocarbon detection and reservoir description. According to the bottleneck problem in the huge calculation, the method which based on genetic algorithm has been proposed in MPD spectral decomposition. Through the experiment of the actual seismic signal, the operation speed which based on the same signal length has increased about 250 times, it can solve large computing problems better.MPD spectrum of the improved decomposition method applied to the reservoir description of the actual work, and has an exploratory research on hydrocarbon detection of thin reservoirs, etc. The studies have shown that MPD improved spectral decomposition method is effective for hydrocarbon detection.