Reliability Analysis Of Suspension Bridge Using Gaussian Process Based Response Surface Method

Wei Zhao,Guo-Shao Su,Li-Hua Hu
DOI: https://doi.org/10.4028/www.scientific.net/AMR.860-863.2970
2014-01-01
Abstract:Aiming to the problems of low precision using traditional response surface method for structural reliability analysis with high nonlinear implicit performance function, Gaussian process regression (GPR) model reconstructing response surface was hybridized into the checking design point method for solving the reliability. Then, an iterative algorithm is presented to reduce the errors of GPR response surface self-adaptively. Thus, a new method namely Gaussian process based response surface for reliability analysis of suspension bridge was proposed. The research results show that the proposed method is feasible. The proposed method has advantages of high efficiency and excellent adaptability for reliability analysis of the complex structural such as suspension bridge.
What problem does this paper attempt to address?