自适应重要抽样方法的改进算法  

IMPROVED ADAPTIVE IMPORTANCE SAMPLING ALGORITHM

作　　者：陈向前[1] 董聪[1] 闫阳[1] 

机构地区：[1]清华大学土木工程学系,北京100084

出　　处：《工程力学》2012年第11期123-128,共6页Engineering Mechanics

基　　金：国家"十一五"科技支撑项目(2006BAJ01B01-2)

摘　　要：失效概率的计算是结构可靠度分析的核心问题之一,发展精确高效的失效概率估计方法渐成国际学术与工程界关注的焦点。该文提出了一种基于样本概率密度加权的采样中心确定方法,该方法兼顾了以下2个目标:1)增加有效抽样中对失效概率贡献大的样本出现的概率;2)提高有效抽样比例。通过将该方法与基于主动引导技术的自适应抽样方法相集成,得到了一种改进的自适应重要抽样方法。理论分析与数值算例表明:该文提出的自适应重要抽样算法具有精度高、计算量小的优点。Structural reliability analysis requires an accurate and efficient evaluation for failure probability. A new scheme is proposed to identify the optimal importance sampling center, based on weighing samples by their probability density functions （PDF）. By incorporating the scheme into the adaptive importance sampling method ISAG （adaptive Importance Sampling based on Active Guiding technology）, an improved adaptive importance sampling method is proposed. The accuracy and efficiency of the proposed method is demonstrated and verified by two numerical experiments.

关 键 词：结构可靠度 自适应 重要抽样 最优抽样中心 主动引导 

分 类 号：TB114.3[理学—概率论与数理统计]