Nonlinear Inversion Of The Residual Static Correction Theoretical Research

This thesis discusses the basic theory of static correction in brief,and the methodof liner residual static correction and nonlinear residual static correction in detail. Weput forward an improved approach based on the advantage and the disadvantage of theabove method respectively. We also introduce the soft design and development for thissystem.Static Correction is the primary step in Seismic Data processing. Statics includeyield statics and residual statics. Residual statics that uses reflection information isprocessed after yield statics. The conventional residual staics are all linear. When thesurface of surveys isn't complex, and the static correction problem isn't obvious, theconventional methods have good result. But, when the surface of surveys is complex,and the statics problem is obvious, the conventional methods can't resolve the staticsproblem, because the linear approaches can converge at the local-optimization result,and result in cycle skips. On the contrary, the nonlinear inversion can obtain theglobal optimization estimations, and avoid cycle skips. In this paper, I introduceseveral nonlinear residual statics methods: stack power methods, simulated annealing,and genetic algorithms and raise a mix approach of linear and unlinear methods.Residual statics processing is only a minor aspect of seismic data processing, sowe can develop this program in blocking style. This system (Seisoft) is developedbased on object-oriented programming technique and the platform of Windows byusing VC.net developing tool. So the system has characteristics of simpleness,flexibility and convenience.The result shows that most of the problem of large residual statics' value can beresolved by using nonlinear residual statics method. In case of resolving largestatics' value, we can consider settling this problem by using nonlinear method insteadof custom approach.