Optimal Performance-Based Design of Steel Frames Using Advanced Metaheuristics

In the present paper, seismic performance-based optimization of steel frames is implemented using a number of advanced meta-heuristic algorithms. The optimization process is implemented by two newly developed meta-heuristics, dolphin echolocation optimization (DEO) and enhanced colliding bodies optimization (ECBO). In the present study, a slight modification is achieved on the ECBO to improve its convergence rate and therefore it is named here as ECBO-II. Furthermore, the results of DEO, ECBO and ECBO-II are compared with those of particle swarm optimization (PSO). Evaluating the structural responses by nonlinear time-history analysis during the optimization process is very time consuming. In order to reduce the computational rigor, radial basis function (RBF) neural network is utilized to predict the necessary structural time-history responses during the optimization process. Two numerical examples are presented to illustrate the efficiency of the meta-heuristics for tackling the seismic performance-based optimization problems.