随机与区间不确定性下的序列化多学科可靠性分析

摘要

为进行两种不确定共存情况下的多学科可靠性分析,提出一种基于概率论和凸集理论的随机与区间不确定性下的序列化的多学科可靠性分析方法.该方法通过对耦合循环解耦,组成多学科概率可靠性分析、基于凸模型的多学科非概率可靠性分析和多学科分析的单循环递推迭代过程,并集成全局灵敏度方程和KKT替代条件替换复杂的凸模型极值分析,保证了多学科可靠性分析具有较好的计算效率.通过两个算例验证了该方法的有效性.%To analyze the multidisciplinary reliability with coexistence of two uncertainties,a sequential multidisciplinary reliability analysis method based on probability theory and convex set theory under random and interval uncertainty was proposed.Through decoupling circularly,the reliability analysis for multidisciplinary probability,the reliability analysis for convex-based multidisciplinary non-probability and the single cycle recursive iteration of multidisciplinary analysis were constituted.The global sensitivity equations and KKT conditions were used to replace the extreme value analysis of convex model,thus the computational efficiency of multidisciplinary reliability analysis was ensured.Two applications were demonstrated to verify the effectiveness and efficiency of proposed method.