Risk design optimization using many-objective evolutionary algorithm with application to performance-based wind engineering of tall buildings

Abstract

Risk design optimization (RDO) is a competent approach for automated performance-based structural design by achieving a balance between safety and economy. Performance-based wind engineering (PBWE) is aimed at improving the life-cycle functionality of wind-sensitive structures, hence could be the very field RDO is tailor-made for. In this paper, we embed PBWE of tall buildings into RDO and tackle some difficulties when integrating them directly. We first formulate an integrated stiffness and vibration control RDO problem, and employ a frequency domain closed-form solution for uncertainty quantification and uncertainty propagation through the excitation-response-performance chain. Then we reveal the multi-objective optimization nature of RDO, and circumvent the difficulties in serviceability loss estimation by replacing scalar total cost with high-dimensional objective vector. Micro multi-objective particle swarm optimization in conjunction with kernel-learning based principle component analysis is employed to solve the corresponding many-objective problem with multiple probabilistic constraints and discrete design variables. The optimization results of CAARC benchmark indicate that we simplify risk-based PBWE of tall buildings from a complex multi-objective decision making process into a relatively easy multi-attribute decision making process. Accordingly, convincing decisions can be made based on the explicit building performance rather than the unreliable loss information. © 2014 Elsevier Ltd.