In this paper, a new approach for estimating the probability of failure is proposed. Due to high computational cost of conventional sampling-based reliability analysis methods, this approach sequentially adds new samples at the proper location to incr...
In this paper, a new approach for estimating the probability of failure is proposed. Due to high computational cost of conventional sampling-based reliability analysis methods, this approach sequentially adds new samples at the proper location to increase efficiency. The method consists of two parts; space partition part and additional sampling part. Space partition is conducted using Voronoi diagram, and additional sampling location is determined using H cell strategy which discriminates the subspaces near the limit state function. The efficiency is compared with the results of other sampling-based reliability analysis methods and they demonstrate the efficiency of the proposed method.