Weight optimization of truss structures by a new feasible boundary search technique hybridized with firefly algorithm

In most metaheuristic optimization techniques, constraint handling is accomplished by traditional penalty function method which may generate infeasible solutions. The main objective of this paper is to introduce a new technique for constraint handling in size optimization of truss structures. The method restricts the searching space on the feasible boundary and hence it is referred to as Feasible Boundary Search (FBS) technique. By limiting the searching space on the feasible boundary, the problem constraints are automatically satisfied. The method generates absolutely feasible high-quality solutions. Its generality enables the scheme to be hybridized with most metaheuristics. By presenting some modifications on the standard Firefly Algorithm (FA), a powerful hybridized technique which is denoted by FBSFA is then developed for fast weight optimization of truss structures. Several design examples show the effectiveness, robustness and accuracy of FBSFA.