Performance comparison of metaheuristic algorithms for optimisation of steel frames using nonlinear inelastic analysis.
Abstract
Nowadays, structural optimisation using metaheuristic algorithms has been widely used because of the significant benefit of these algorithms in solving highly complicated problems. With the same advantages, many optimisation algorithms have been proposed. However, depending on the characteristics of each optimisation problem type, the algorithm shows different performance. In this paper, the optimisation of steel frames using nonlinear inelastic analysis is considered. The performance of four recent metaheuristic algorithms, including the efficient pbest differential evolutionary (EpDE) algorithm, the enhanced colliding bodies optimisation (ECBO), harmonic search (HS), and Rao algorithms, for this optimisation problem, is compared. A 5x5 planar steel frame is optimised where the objective function is the total weight of the structure and the constraints include both load-carrying capacity and displacement requirements. The results showed that the EpDE algorithm found the best optimal results and had the best converged speed.