基于广义遗传算法的自适应重要抽样理论

1 引言随着计算技术和软件工具的迅猛发展,具有良好普适性的数值模拟方法在许多领域开始受到了广泛的重视。由于Harbitz,Melchers,Wu和本文作者的一系列工作,基于重要抽样理论的非线性系统可靠性分析技术取得了很大的进展。尽管如此,基于重要抽样理论的非线性系统可靠性分析技术中一些带有根本性的理论问题并没有彻底解决。突出表现是,对于广义多设计点问题、寻找所有设计点的求解算法不仅没有提出,甚至没有提出过合理的求解思路。造成这一局面的原因是,多设计点问题的求解需以非连通域非线性系统全局优化问题的解决为前提,而后一问题在

An Adaptive Importance Sampling Theory Based on The Generalized Genetic Algorithm

In the present paper,using the generalized genetic algorithm,the problem of finding out all design points in the case of generalized multiple design point is solved,establishing recursion-type bound-and-classification algorithm,the problem of reducing and synthesizing generaliged multiple design points is also solved.The present paper shows that the adaptive importance sampling theory based on the generalized genetic algorithm is an more efficient tool for the reliability simulation of nonlinear sys-tems.