Genetic Inversion Method Of Two-dimensional Density Interfaces

Gravity research is one of the important geophysical methods in the investigation of the crustal and mantle structure. The information of density interface by the gravity potential field is very important for the research of tectonic structure, regional geologies, metal and nonmetal exploration. In particular, the inversion of this information can not be replaced for deep crust. However, decomposition of the potential field with multiple source (different depth , scale and properties of matter) is very difficult in exploration geophysics.Genetic algorithms (GA) have been developed in recent years. GA elements mimic biological processes that have been studied in natural evolution, namely that of natural selection and survival of the fittest. GA provide computational procedures that find the optimization solution. GA are a better global and local optimization method. Based on its characteristic GA have become an effective inversion method for the research of the geophysical potential field.The 2-D surface model of geologic structures on Marqen-Jingbian profile is established by forward-inverse method. Its Moho depth are calculated by GA inversion method. The example are presented in the paper, the results show to be ideal, which prove it available for GA to inverse gravity data caused by 2-D surface of geologic structures. The gravity anomalies are analyzed through 2-D surface density model. Combined with other regional geological and geophysical information, exploratory research of the crust structure and dynamic mechanism is done finally.