A Differential Evolution Based Algorithm for Constrained Multiobjective Structural Optimization Problems


Abstract

Structural optimization problems aim at increasing the performance of the structure while decreasing its costs guaranteeing, however, the applicable safety requirements. As these aspects are conflicting, the formulation of the structural optimization problem as multiobjective is natural but uncommon, and has the advantage of presenting a diverse set of solutions to the decision maker(s). The literature shows that Evolutionary Algorithms (EAs) are effective to obtain solutions in multiobjective optimization problems, and that the Differential Evolution (DE) based ones are efficient when solving structural mono-objective structural optimization problems, specially those with a real encoding of the design variables. On the other hand, one can note that DE has not yet been applied to the multiobjective version of these problems. This article presents a performance analysis of a DE-based algorithm in five multiobjective structural optimization problems. The obtained results are compared to those found in the literature, and the comparisons indicate the potential of the proposed algorithm.