Seismic Wavelet Estimation Based On High-order Statistics And Chaotic Genetic Algorithm

As the basis of deconvolution, wave impedance inversion and forward model, the seismic wavelet estimation is a long-term research in seismic data processing. The quality of its recovery affects the seismic signal processing and data explanation directly. Conventional wavelet extraction methods usually have assumption that the wavelet is in certain form. Based on the MA model hypothesis of the seismic trace, the cumulant matching method is proposed in this thesis. This method not only retained the phase information of the seismic data , but also effectively eliminated the Gaussian noise.The cumulant matching method was a nonlinear inversion question essentially, and solving the target function was a highly nonlinear optimization problem. This thesis detailedly studied Genetic Algorithms and Chaotic optimization method, then combined them to propose a Chaotic Genetic Algorithm (CGA) for seismic wavelet calculation. The Chaotic Genetic Algorithm effectively maitains population diversity and avoid premature phenomenon in the evolution. Several typical benchmark functions numerical simulations show that CGA is an effective nonlinear inversion algorithm with fast global convergency and high stability. The numerical simulations and real seismic data processing demonstrate the application and stability of the wavelet estimation method based on fourth-order matching and CGA.The precision and efficiency of the seismic wavelet extraction gain obvious improvement with the application of the new method. The achievements indicate the broad application future of high-order statistics in the field of the seismic data processing.