Structural shape optimization of free-form surface shell and property of solution search using firefly algorithm

Abstract

The computation procedure for structural shape optimization is based on the heuristic optimization procedures, such as swarm intelligence (SI), that use patterns found in self-organizing phenomena observed in nature. Most SI techniques including particle swarm optimization (PSO) and artificial bee colony (ABC), attain a global optimal solution. The firefly algorithm (FA) can attain both a global optimal and local optimal solutions by setting suitable computational parameters. However, the method for setting these parameters is comparatively difficult. In order to simplify the setting of these parameters, we implement a computational scheme where the distance between two fireflies in the design variable space is dimensionless. The effectiveness and the validity of this FA are clarified by showing the diversified solution for a global optimal solution and local optimal solutions using a local search in the structural shape optimization of a free-form surface shell.