Study On Several Issues Of Deconvolution And Wavelet Extraction For Seismic Signal

Seismic wavelet is widely used in the field of seismic exploration. Seismic waveletis not only used in high-resolution deconvolution of seismic data, forward modeling andthe wave impedance inversion, but also used in horizon calibration, matching of loggingdata and seismic data. Seismic wavelet is the link between logging data and seismic data.Improvement and deepening of the seismic wavelet extraction methods is a majorresearch content and key technology in the field of seismic signal processing. Forinstance, the quality of inversion is closely related to the accuracy of seismic waveletextraction. In addition, the seismic signal deconvolution is also one of the importantmeans to improve the seismic resolution. Therefore, both the seismic wavelet extractionand deconvolution methods are important basis for seismic interpretation and reservoirprediction. The major work in this thesis as follows:(1)Study on the basic theory of seismic signal deconvolution, including seismicconvolution model, least squares deconvolution, predictive deconvolution andhomomorphic deconvolution methods and analyzes the existing problems anddisadvantages of these methods. Finally, the actual seismic data are used in thesimulation experiments of these methods and the results were analyzed.(2) Study on the basic theory and method of seismic wavelet extraction. Mainlystudy on second-order statistics wavelet extraction, RoyWhite wavelet extraction andhigher order statistics wavelet extraction methods. Finally, the actual seismic data areused in the simulation experiments of these methods and the results were analyzed.(3)Present a new seismic wavelet extraction method that the optimal filteringtheory in fractional domain is applied to seismic wavelet extraction. Research onoptimal filtering in fractional domain and the basic theory and principles of this newwavelet extraction algorithm. Finally, the theoretical seismic data are used in thesimulation experiments of this new method and the results of the new method are betterthan the results of the method in traditional Fourier domain.(4)Core algorithms with C++language were realized and software modules of theseismic wavelet extraction and seismic deconvolution on QT development platform based on Windows or Unix operating system are developed. Finally, we test thesoftware modules with the theoretical data and the actual data, and prove the correctnessand practicability of methods introduced in this thesis and software modules wedeveloped.