2015 年 64 巻 3 号 p. 190-195
The purpose of this paper is to propose a method effective for the structural reliability assessment based on the multimodal analysis. A structure should be evaluated from various perspectives in order to improve the safety against unpredictable crises. Factors related to failure modes and elements contributing to the capability are estimated based on each criterion. Thus, by considering these criteria as modes, the multimodal analysis can extract information effective to the improvement of reliability and the rational preparedness. However, because this analysis requires enormous samples in practical problems, it is difficult to apply sampling methods like Monte Carlo simulation. In this paper, an attempt is made to propose a search method of samples feasible to the multimodal analysis by using Cellular Automaton Particle Swarm Optimization. A numerical example with a single-bay and single-story rigid frame structure is presented to demonstrate the applicability of the proposed method.