Reliability-based Optimization of Spatial Structures using Aproximation Model

Abstract

This study presents an optimization approach for the reliability-based design of spatial structures subjected to severe earthquakes, while uncertainty involved in structural analysis parameters is taken into consideration. Simulated annealing method is utilized to find the structure with minimum weight from available sections, while the specified structural performances in terms of probability are satisfied. The dynamic performance of a spatial structure with uncertainty is evaluated in a probabilistic manner by applying Monte Carlo simulation (MCS). To reduce the high computational costs of MCS, Kriging method is adopted to generate the approximate response surface.