Using genetic algorithm to determine 2-D gravity modeling of sedimentary basins with density contrast varying parabolically with depth

Abstract

A program of genetic algorithm has been developed to estimate the depth of a 2-D sedimentary basin whose density contrast varies with depth according to a parabolic law. The model was built consisting of 2-D vertical juxtaposed prisms. Depths of the prisms, computed by genetic algorithm based on random values and optimal depths were finally found after many generations of evolution. The genetic algorithm using the fitness function was combined by root mean square error of data and "norm" model and the latter was multiplied by a Tikhonov regularization parameter to stabilize the solutions. Firstly, the method was tested on a model and its result were coincident with the model. Secondly, it was applied to interprete a profile of gravity anomaly in Mekong Delta. The results showed that the calculate and observed gravity anomalies were well fitted.