Multi-objective optimization of grillages using genetic algorithm with parallel strategy

Abstract

The paper proposes a technology based on simple weighting technique for multi-objective optimization of grillagetype foundations with genetic algorithm. Original program, written by FORTRAN programming language, is applied for the optimization of 10-pile and 15-pile grillages. The set of design parameters contains the pile positions and the cross-sectional dimensions of linking beams. While genetic algorithm usually exhibit slow convergence rate after a particular number of populations, the parallel strategy is proposed to achieve better optimization results. The outcome of numerical experiments proved parallel strategy to be efficient method for the solution.