This paper deals with an adaptation of artificial neural networks in the context of the reliability analysis of non-linear limit state functions. An extreme learning machine (ELM) that is categorized as a single-hidden-layer feed-forward neural network is considered in the present study. Using a trained ELM-based approximate meta-model, the reliability analysis is conducted in conjunction with Monte Carlo simulation. The ELM is compared with both single and multiple-hidden-layer back-propagation neural networks. A number of non-linear and large-dimensionality limit state functions are explored to support the proposed method in terms of approximation accuracy and reliability index.
SongCYLeeJ. Reliability based design optimization of knuckle component using conservative method of moving least squares meta-models. Prob Engng Mech2011; 26(2): 364–379.
2.
DimouCKKoumousisVK. Competitive genetic algorithms with application to reliability optimal design. Adv Engng Softw2003; 34: 773–785.
3.
LagarosNDPapadopoulosV. Optimum design of shell structures with random geometric, material and thickness imperfections. Int J Solids Struct2006; 43: 6948–6964.
PapadrakakisMLagarosND. Reliability-based structural optimization using neural networks and Monte Carlo simulation. Comput Methods Appl Mech Engng2002; 191: 3491–3507.
6.
ElhewyAHMesbahiEPuY. Reliability analysis of structures using neural network method. Prob Engng Mech2006; 21: 44–53.
7.
BucherCMostT. A comparison of approximate response functions in structural reliability analysis. Prob Engng Mech2008; 23(2-3): 154–163.
8.
HuangG-BZhuQ-YSiewC-K. Extreme learning machine: theory and applications. Neurocomputing2006; 70: 489–501.
9.
SerreD. Matrices: theory and applications. New York: Springer, 2002.
10.
GondalZUR. Reliability assessment using neural network based approximations and Monte Carlo simulation. M.S. Thesis, Yonsei University, Korea, 2008.
11.
LeeJJeongHChoiD-H. An enhancement of constraint feasibility in BPN based approximate optimization. Comput Methods Appl Mech Engng2007; 196: 2147–2160.
12.
LeeJKangS. GA based meta-modeling of BPN architecture for constrained approximate optimization. Int J Solids Struct2007; 44(18-19): 5980–5993.
13.
LeeJJeongHKangS. Derivative and GA based methods in metamodeling of back-propagation neural networks for constrained approximate optimization. Struct Multidiscip Optim2008; 35(1): 29–40.
14.
PichenyVKimNHHaftkaRT. Application of bootstrap method in conservative estimation of reliability with limited sample. Struct Multidiscip Optim2010; 41: 205–217.
