Surrogate-model-based reliability method for structural systems with dependent truncated random variables
Surrogate-model-based reliability method for structural systems with dependent truncated random variables
复制标题
具有相关截断随机变量的结构系统基于代理模型的可靠性方法
DOI：
10.1177/1748006x17698065
复制
发表时间：
2017-03
期刊：
影响因子：
--
通讯作者：
Zhangchun Tang
中科院分区：
文献类型：
--
作者：
Ning-Cong Xiao;Libin Duan;Zhangchun Tang
有效计算具有相关截断随机变量和多种失效模式的结构系统的失效概率和可靠性敏感性是一个挑战，主要是由于多种失效模式的复杂特征和交叉点以及相关的性能函数。在本文中，针对具有相关截断随机变量和多种失效模式的结构系统，提出了一种新的基于代理模型的可靠性方法。 Copula 函数用于对截断随机变量的相关性进行建模。在支持的区间内生成小尺寸的均匀分布样本，以完全正确地覆盖整个不确定性空间。基于所提出的训练点和支持向量机构建了精确的代理模型，以在几乎整个不确定性空间中准确地近似输入和系统响应之间的关系。推导了基于所构建的替代模型计算具有截断随机变量和多种失效模式的结构系统的失效概率和可靠性敏感性的方法。使用两个数值例子证明了所提出方法的准确性和效率。
Calculating probability of failure and reliability sensitivity for a structural system with dependent truncated random variables and multiple failure modes efficiently is a challenge mainly due to the complicated features and intersections for the multiple failure modes, as well as the correlated performance functions. In this article, a new surrogate-model-based reliability method is proposed for structural systems with dependent truncated random variables and multiple failure modes. Copula functions are used to model the correlation for truncated random variables. A small size of uniformly distribution samples in the supported intervals is generated to cover the entire uncertainty space fully and properly. An accurate surrogate model is constructed based on the proposed training points and support vector machines to approximate the relationships between the inputs and system responses accurately for almost the entire uncertainty space. The approaches to calculate probability of failure and reliability sensitivity for structural systems with truncated random variables and multiple failure modes based on the constructed surrogate model are derived. The accuracy and efficiency of the proposed method are demonstrated using two numerical examples.
登录
查看更多内容
DOI：
10.1007/978-1-4899-7687-1_810
发表时间：
2017
期刊：
--
影响因子：
--
作者：
Xinhua Zhang
通讯作者：
Xinhua Zhang
DOI：
10.1201/9781003139041-11
发表时间：
2021-03
期刊：
An Introduction to IoT Analytics
影响因子：
--
作者：
Harry G. Perros
通讯作者：
Harry G. Perros
影响因子：
2.6
作者：
Kaymaz, I;McMahon, CA
通讯作者：
McMahon, CA
影响因子：
7.4
作者：
Ke-Lin Du;M. Swamy
通讯作者：
Ke-Lin Du;M. Swamy
影响因子：
8.1
作者：
Bichon, Barron J.;McFarland, John M.;Mahadevan, Sankaran
通讯作者：
Mahadevan, Sankaran