Optimal truss sizing by modified Rao algorithm combined with feasible boundary search method

Abstract

Newly developed optimization algorithms, the Rao algorithms, are utilized for optimal truss sizing in this study. In particular, a modified Rao algorithm that takes advantage of Rao-1and Rao-2 algorithms is proposed. The modified algorithm retains the main feature of Rao algorithms, i.e., simplicity and parameter-freeness. The modified Rao algorithm is then hybridized with the Feasible Boundary Search (FBS) technique to form a constrained Rao algorithm, named FBSmRao. The applicability of the original Rao algorithms, including Rao-1 and Rao-2, and the proposed FBSmRao for truss sizing optimization is validated through four benchmark design problems with continuous variables. The results show that Rao algorithms can be used to obtain optimum weight for trusses. Moreover, the proposed FBSmRao outperforms the original Rao algorithms in terms of solution quality and convergence speed. Comparing with some other metaheuristics, including ones using FBS, FBSmRao can obtain competitive designs, especially for large-scale structures.