Hybrid Global Algorithm Wavelet Estimation And Extraction Via Higher Order Cumulants

The novel method of Estimation and extraction of seismic Wavelet-Higher order cumulants is presented in this thesis at first. Target function is highly nonlinear optimization problem in this method. When using singer method -simulated annealing, it takes a large number of calculations and a great deal of runtime, and it difficult to get precision solution. So, simulated annealing and simplex combine to form a global hybrid method. Numerical simulation result shows that the hybrid method is a high effective hybrid global algorithm. It is of high performance in nonlinear optimization of high dimension complex function. And this method is applied in estimation and extraction of seismic Wavelet via higher order cumulants. Theory model and real data test of this method show that it is of fast constringency and of fairly robust to initial conditions.