基于原始-对偶内点法寻优的波阻抗稀疏反演方法

ABSTRACTABSTRACTThe deconvolution of seismic data and impedance inversion are keytechnologies to improve the resolution,reservoir prediction and discrimination.Impedance is the physical property of the reservoir,and reflectivities is the sequencewhich derived from the different impedances of reservoirs that can be processed bydeconvolution of seismic data and the known wavelet.Above is the basic of thispaper.To improve the resolution and speed up the computing,a new method ofdeconvolution is proposed,which based on the Primal-Dual method for convexoptimization.The object function minimizes the l norm of the reflectivities whiletrying to keep the convolution of them the same as the original data with some bias.With the Primal-Dual method,the minimization can be solved as an unconstrainedminimization,which is part of the convex optimization.Two parts constitute thesolving,one is the outer iterations,called Sequential Unconstrained MinimizationTechnique(SUMT),stopping criterion of which is the Duality Gap;the other is theinner iterations looking for minimization under certain Duality Gap using NewtonMethod.The method this paper proposed has improved the resolution,while keeping theSNR falling down little enough.Linear Equations need to be solved in the solving,and LSQR is applied to so memory,CPU time can be saved,and high accuracy,quick convergence can be reached.KEYWORDS:impedance inversion,interior-point method,pre-conditioning,LSORv