A cumulative belief-degree approach for nuclear safeguards evaluation

Nuclear safeguards are a set of activities to verify that a State is living up to its international undertakings not to use nuclear programs for nuclear weapons purposes. International Atomic Energy Agency (IAEA) uses a hierarchical assessment system that is composed of critical activities in the nuclear fuel cycle, processes required for the activities, and the indicators to evaluate the processes. IAEA experts benefit from several sources to evaluate the indicators such as State declarations, on-site inspections, IAEA databases, and open sources. One of the most important problems in nuclear safeguards evaluation (NSE) is observed in the aggregation of the multiple expert evaluations. In this study, a methodology is proposed to solve this problem where fuzzy linguistic terms are proposed to represent the expert evaluations and cumulative belief degrees (CBDs) are applied to aggregate the evaluations. CBD is introduced based on belief structure, which is developed on the basis of decision theory and Dempster-Shafer theory of evidence. The study also presents a numerical example to show the applicability of the proposed methodology.