ABSTRACT
P. Baraldi & E. Zio Department of Energy, Polytechnic of Milan, Milan, Italy
ABSTRACT: This work considers a situation in which multiple assumptions are made on the behaviour of the system under analysis, giving rise to corresponding alternative plausible modeling hypotheses. The opinions of different experts are sought on the correctness of the hypotheses of the individual assumptions. It is assumed that the experts do not have sufficiently refined knowledge to probabilistically characterize the uncertainty in the correctness of the different hypotheses; in alternative, the belief and plausibility functions of the DempsterShafer theory are used for uncertainty representation. Then, the Dempster rule of combination is used for the aggregation of the uncertainties of different experts on the correctness of the hypotheses and the propagation of the uncertainties regarding the individual assumptions onto the uncertainty characterizing the entire model. A case study is presented, with regards to the assumptions underpinning a model of groundwater flow for use in the performance assessment of radioactive waste repositories.
