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

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.

• Periodical: Advanced Materials Research (Volumes 860-863)
• Pages: 2970-2974
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.860-863.2970
• Online since: December 2013
• Authors: Wei Zhao, Guo Shao Su*, Li Hua Hu
• Keywords: Gaussian Process, Implicit Limit State, Response Surface Method (RSM), Structural Reliability

References

[1] G. F Zhao: Engineering Structure Reliability Theory and Application. (Dalian University of Technology Press, China 1987).

[2] B. Kims, M. Nas: Str. Saf. Vol. 19. (1997), p.3.

[3] B. Chriatain, M.A. Thomas: Appl. Math. Mech. Vol. 23. (2008), p.154.

[4] J. Deng, D.H. Gu, X.B. Li, Z.Q. Yue: Str. Saf, Vol. 27. (2005), p.25.

[5] C. Jin, Q.S. Li, C.X. Ru: Pro. Eng. Mec, Vol. 23. (2008), p.51.

[6] J.S. Zhu, R.C. Xiao, L.Z. He : China Civil Engineering Journal. Vol. 40. (2007), p.41.

[7] J. E. Hurtado: Str. Saf. Vol. 26. (2004), p.271.

[8] M. Seeger: Int. Jou. Neu. Sys. Vol. 14. (2004), p.69.

[9] M. Giromali, S. Rogers: Neural Computation. Vol. 18. (2006), p.1790.