Imprecise subset simulation

The objective of this work is to quantify the uncertainty in probability failure estimates resulting from incomplete knowledge distributions for input random variables. We propose a framework that couples widely used Subset simulation (SuS) with Bayesian/information theoretic multi-model inference. process starts data infer model inputs. Often such sets are small. Multi-model inference assess associated model-form and parameters these variables form probabilities joint parameter densities. A sampling procedure construct set equally probable candidate an optimal importance distribution determined analytically set. then performed using density conditional re-weighted sampling. result empirical provide direct on small sets. method demonstrated be both computationally efficient – requiring only single subset nominal cost sample re-weighting reasonable probabilities.