Uncertainty Assessment Methodology for Probabilistic Risk Assessment (PRA); Data, Methods, Models, and Inputs

摘要

Uncertainty analysis is a crucial step in process of probabilistic risk assessment (PRA) for better management and decision making purposes. This paper reviews the process of uncertainty analysis and methodologies for characterization of the uncertainties and their treatment in probabilistic risk assessment (PRA). This research is limited to Fault Tree (FT) and Event Tree (ET) methodologies only and deals with all uncertainties in process of PRA level I. A literature review was conducted on the subject to evaluate the state of the art on the topic. Uncertainty taxonomy is reviewed in this research to better address different sources of uncertainty. A hybrid method of maximum Entropy approach supported by Bayesian Updating is proposed to quantify the parameters' uncertainties effectively by using all relative and partially relative data and information. Bayesian approach is utilized for the inference of the parameter uncertainties. Examples from applications are provided for greater clarification of the proposed uncertainty analysis techniques.