The Bayes Decision Model And Its Realization In Nuclear Accident Emergency

Nuclear accidents emergency is a complicated system involving all kinds of factors or attributes. Only under the guideline of scientific decision theories and methods and by the aid of computers as real-time decision-making tool, the need of nuclear accidents emergency can be fully satisfied. In so doing, our country has imported Real-time On-line Decision Support System (RODOS) from European Communities which includes three parts: Analysing Subsystem(ASY) N Countermeasure Subsystem(CSY^ Evaluation Subsystem(ESY), and among them The ESY is a gap .S.French has formulated that Bayes analysis method can be used to solve the problem of uncertainty in the decision-making of nuclear accidents. However, up to now, no article on the topic has been published around the world to discuss the concrete application of Bayes method in ESY.The National Project Committee take "Decision System of Nuclear Accidents Emergency(DSNAE)" as a key item.One of the item's aim is:On the base of absorbing RODOS's fruit, research and develop a Decision System of Nuclear Accidents Emergency adapting to our country.One of the content we undertake is using Bayes decision method in ESY.Through the analysis and comparison, we find that the uncertainty of wind field has immense effects on the nuclear accidents emergency decision. In the light of that, after combining with the relative knowledge of nuclear accidents emergency, we put forth the Bayes decision model (NBDM) on the basis of randomicity decision and Bayes risk decision theory in order to decrease the uncertainty of wind field and to improve the precision of decision in ESY.To reflect the preference of experts correctly when valuing schemes, we use the attribute-coordinates analysis method as a tool to establish the geometry so as to help experts make correct preference by producing the attribute preference. And then we establish the lose function L based on these. At last, we realize the NBDM model in computer. At the same time, in order to follow the development of European Community and draw a lesson from their experience, we put the idea of utility function into practice partly, which is put forth by Russia RODOS team. And we compare it with our lose function L which is built by the attribue-coordinates analysis method, we can draw a tentative conclusion that the latter is better than former in time but they can achieve the same effect in elicitation of the preference of experts .