SVM-based Reliability Analysis Method
The reliability analysis of long-span cable-stayed bridges is a challenge for researchers because the fimit state function is usually implicit and the structural analysis by finite element method is complex. The response surface method (RSM) is a popular method to approximate the explicit limit state function. But RSM is not accurate when the limit state function can not be expressed accurately by quadratic polynomial. From the statistical viewpoint this functional approximation corresponds to a regression approach. The Support Vector Machine (SVM) is a good choice for this aim. Once the functional approximation is finished, the common reliability methods, such as the first-order reliability method (FORM), the second-order reliability method (SORM) and Monte Carlo simulation method (MCSM), can be used to calculate the failure probability of the structural systems. The reliability methods based on SVM were developed and used to resolve several benchmark problems and the reliability analysis of a long-span cable-stayed bridge. It shows that the SVM-based reliability method is a powerful tool to handle the complex structural reliability problems.
support vector machines reliability FORM SORM Monte Carlo simulation cable-stayed bridge
Li Wei Yu Xiaolin
School of Civil Engineering and Transportation South China University of Technology Guangzhou, P.R. China
国际会议
The 10th International Conference on Intelligent Technologies(第十届智慧科技国际会议 InTech09)
桂林
英文
584-588
2009-12-12（万方平台首次上网日期，不代表论文的发表时间）