An evaluation method on the integrated safeguards based on fuzzy theory
Introduction
Many countries concluding comprehensive safeguards agreements (INFCIRC/153 type) with the International Atomic Energy Agency (IAEA) are currently going to agree a new additional protocol (INFCIRC/540 type). The protocol will make them provide more information to the IAEA. The expanded framework is expected to establish stronger and more efficient IAEA safeguards system, which we call “Integrated Safeguards”. (See the paragraphs 2 and 23 of [1].) However, there seems to be no evaluation method that enables inspectorate to derive a final evaluation from the information collected through the Integrated Safeguards Implementation. In this paper, the authors try to make the fundamentals clear and propose a method to evaluate the effectiveness of Integrated Safeguards Implementation. In addition, the method is shown to be useful for the planning to improve the efficiency.
Section snippets
Objective of the evaluation
Objective of the comprehensive safeguards is described in INFCIRC/153. Evaluation of the safeguards-effectiveness should be to estimate a degree of assurance to what extent the objective is attained. In this case, the assurance-degree that “no nuclear material in a country is used for manufacture of nuclear explosive devices” should be estimated.
Definition of nuclear material types
Manufacture of nuclear explosive devices will need many steps. However, the final step before weaponisation will be a step to acquire “Highly enriched
Representation by “Fuzzy linguistic variables”
In case of semi-quantitative evaluation, “Fuzzy linguistic variables” will be useful because they can represent human judgement of evaluation naturally. In the application here, they will represent different levels of assurance-degrees. Concretely, the authors propose the definitions of five fuzzy linguistic variables; “nearly one”, “high”, “medium”, “low” and “completely zero”. They are defined by their corresponding membership functions on the interval of [0,1]. Their forms are depicted in
Considerations
(1) Especially in the fields of fuzzy logic applications, various kinds of definitions similar to our definitions of “” and “” have been proposed. Many of them are divided into three groups: “t-norms”, “t-conorms” and “mean operators”. Among them, “t-norms” are often used to represent the meaning of “and” in various applications. But, for example, when we assume that the input–assurance-degrees are 0.6 and 1, the output is 0.6 according to the definitions of “t-norms” (see [4]). On the
Conclusions
(1) Effectiveness of safeguards implementation can be evaluated by the method proposed in this paper. Its evaluation logic trees are given by Fig. 1, Fig. 2. Definitions of “” and “” needed to calculate the logic trees are given by Section 2.2.
(2) From the case-studies using the above evaluation method, implementation of integrated safeguards only to HU and PU is shown to be able to attain high efficiency. As the consequence, the conventional safeguards activities relevant to LU, SF and FF
References (4)
- IAEA General Conference, Strengthening the effectiveness and improving the efficiency of the safeguards system and...
- et al.