Optimization applied to dynamic poroelasticity using boundary element method (BEM) and genetic algorithm (GA)

This paper focused on developing a methodology based on coupling the Genetic Algorithm (GA) and the Boundary Element Method (BEM) for predicting the mechanical properties of a soil if a previous displacement x frequency curve is known. The NSGA-II (Elitist Non-Dominated Sorting Genetic Algorithm) was chosen for optimizing a dynamic poroelastic problem. The revisited problem introduced by Cheng is the benchmark used to verify the numerical routine and to apply the proposed optimization procedure. The present methodology was shown to be able to predict the mechanical properties of the underground soil if a displacement x frequency curve is previously known.