Markov Chain Monte Carlo Simulation Methods for Structural Reliability Analysis. (2017)

• Record Type: Journal Article
• Authors: Proppe, Carsten
• Is Part Of: Procedia engineering. Volume 199(2017)
• Page Start: 1122
• Page End: 1127
• Publication Date: 2017
• Subjects: Reliability Estimation -- Markov Chain Monte Carlo Simulation -- Subset Simulation -- Moving Particles Algorithm; Engineering -- Congresses; Engineering -- Periodicals; Engineering; Conference proceedings; Periodicals; 620.005
• DOI: 10.1016/j.proeng.2017.09.226
• Languages: English
• ISSNs: 1877-7058

Abstract

Abstract: Two Markov chain Monte Carlo simulation methods for reliability estimation, subset simulation and the moving particles algorithm, are compared. To this end, both low-dimensional and high-dimensional test cases are considered. The investigation sheds light on the avoidance of correlated samples in both algorithms, notably on the necessity of a burn-in period and the influence of a metamodel.