Efficiency Improvement of Stochastic Simulations by Means of Subset Sampling
M. Liebscher, S. Pannier, W. Graf (TU Dresden) The design of engineering systems requires a sophisticated structural analysis close to reality. The uncertainty of structural parameters, such as loads, material parameters, and geometrical properties must be taken into account. Based on the framework of probabilistics a structural behavior may be assessed by a failure probability. The failure probability can be computed with the aid of Monte Carlo simulation. Especially in the case of complex nonlinear structures the Monte Carlo simulation meets their limits. The more advanced simulation method subset sampling, which promises to compensate this drawback, is investigated in this paper. The main idea of subset sampling is the subdivision of the failure event into a sequence of partial failure events, which are denoted as subsets. The numerical efficient sampling within the subset is realized with aid of the Markov chain Monte Carlo simulation. Subset sampling is demonstrated by means of examples. Keywords: subset sampling, Monte Carlo simulation, reliability analysis
https://www.dynamore.de/en/downloads/papers/06-forum/papers/robustness-optimization/efficiency-improvement-of-stochastic-simulations/view
https://www.dynamore.de/@@site-logo/DYNAmore_Logo_Ansys.svg
Efficiency Improvement of Stochastic Simulations by Means of Subset Sampling
M. Liebscher, S. Pannier, W. Graf (TU Dresden) The design of engineering systems requires a sophisticated structural analysis close to reality. The uncertainty of structural parameters, such as loads, material parameters, and geometrical properties must be taken into account. Based on the framework of probabilistics a structural behavior may be assessed by a failure probability. The failure probability can be computed with the aid of Monte Carlo simulation. Especially in the case of complex nonlinear structures the Monte Carlo simulation meets their limits. The more advanced simulation method subset sampling, which promises to compensate this drawback, is investigated in this paper. The main idea of subset sampling is the subdivision of the failure event into a sequence of partial failure events, which are denoted as subsets. The numerical efficient sampling within the subset is realized with aid of the Markov chain Monte Carlo simulation. Subset sampling is demonstrated by means of examples. Keywords: subset sampling, Monte Carlo simulation, reliability analysis
K-I-4.pdf
— 1.6 MB