期刊
PROBABILISTIC ENGINEERING MECHANICS
卷 63, 期 -, 页码 -出版社
ELSEVIER SCI LTD
DOI: 10.1016/j.probengmech.2021.103119
关键词
High-dimensional stochastic dynamical systems; First-passage reliability; Equivalent drift coefficient; Absorbing boundary condition; Ensemble-evolving-based generalized density evolution equation (EV-GDEE) br; Ensemble-evolving-based generalized density evolution equation (EV-GDEE)
资金
- National Natural Science Foundation of China (the National Distinguished Youth Fund of NSFC) [51725804]
- National Natural Science Foundation of China (NSFC) [11672209, 51538010]
- National Natural Science Foundation of China (NSFC-Guangdong Province Joint Project, PR China) [U1711264]
- NSFC-DFG joint program [11761131014]
- International Joint Research Program of Shanghai Municipal Government, PR China [18160712800]
A novel method is proposed in this paper to capture the time-variant first-passage reliability of high-dimensional systems by imposing appropriate absorbing boundary condition on EV-GDEE and using a feasible numerical method, achieving higher efficiency and accuracy.
The reliability analysis of high-dimensional stochastic dynamical systems subjected to random excitations has long been one of the major challenges in civil and various engineering fields. Despite great efforts, no satisfactory method with high efficiency and accuracy has been available as yet for high-dimensional systems even when they are linear systems, not to mention generic nonlinear systems. In the present paper, a novel method by imposing appropriate absorbing boundary condition on the newly developed ensemble evolving-based generalized density evolution equation (EV-GDEE) combined with a feasible numerical method is proposed to capture the time-variant first-passage reliability of high-dimensional systems enforced by additive white noise excitation. In the proposed method, the equivalent drift coefficients in EV-GDEE can be estimated by analytical expression or captured by some representative deterministic dynamic analyses. Further, imposing the absorbing boundary condition and then solving the EV-GDEE, a one-or two-dimensional partial differential equation (PDE), yield the remaining probability density of the response of interest. Consequently, by integrating the remaining probability density, the numerical solution of time-variant first-passage reliability can be obtained. Several numerical examples are illustrated to verify the efficiency and accuracy of the proposed method. Compared to the Monte Carlo simulation, the proposed method is of much higher efficiency. Problems to be further studied are finally discussed. ABSTRACT The reliability analysis of high-dimensional stochastic dynamical systems subjected to random excitations has long been one of the major challenges in civil and various engineering fields. Despite great efforts, no satisfactory method with high efficiency and accuracy has been available as yet for high-dimensional systems even when they are linear systems, not to mention generic nonlinear systems. In the present paper, a novel method by imposing appropriate absorbing boundary condition on the newly developed ensemble evolving-based generalized density evolution equation (EV-GDEE) combined with a feasible numerical method is proposed to capture the time-variant first-passage reliability of high-dimensional systems enforced by additive white noise excitation. In the proposed method, the equivalent drift coefficients in EV-GDEE can be estimated by analytical expression or captured by some representative deterministic dynamic analyses. Further, imposing the absorbing boundary condition and then solving the EV-GDEE, a one-or two-dimensional partial differential equation (PDE), yield the remaining probability density of the response of interest. Consequently, by integrating the remaining probability density, the numerical solution of time-variant first-passage reliability can be obtained. Several numerical examples are illustrated to verify the efficiency and accuracy of the proposed method. Compared to the Monte Carlo simulation, the proposed method is of much higher efficiency. Problems to be further studied are finally discussed.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据