An improved multi-objective optimization approach for performance-based design of structures using nonlinear time-history analyses

Performance-based design (PBD) of buildings can be properly addressed in a multi-objective optimization framework. However, computational costs of such an approach will be very expensive especially if nonlinear time-history analysis (NTHA) is used as the evaluation tool. In this paper, significant reductions in computational costs of solving structural multi-objective optimization problems is achieved by proposing a new metaheuristic surrogate model called Surrogate FC-MOPSO. In this method, which is an extension of FC-MOPSO algorithm, NTHA and pushover analysis (PA) are simultaneously employed for evaluating system responses. PAs are adopted as an approximation tool in the surrogate model while the responses corresponding to feasible solutions are always evaluated from NTHAs. The final Pareto optimal solutions, which yield tradeoffs between initial and life-cycle costs (LCCs), are already evaluated based on NTHAs. It is shown that application of the proposed method results in substantial reductions of runtime of the considered problems. It is also demonstrated that adopting PAs as the only evaluation tool in optimal performance-based design of structures can result in unreliable solutions.