Estimation of fuel concentration distribution and subcriticality by using “Pattern matching” method

Development of criticality safety monitoring is an important issue in nuclear fuel reprocessing plants. In this paper two subcriticality estimation methods are compared. One method is based on conventional arti®cial neural network method. But its performance was no good in the sense that the accuracy of estimated subcriticality was the same order as the results estimated by the conventional Modi®ed Neutron Source Multiplication method. The other method considered is referred to as the ``Pattern matchingʹʹ method. It can be used as the critical safety monitor for non-uniform fuel solution systems in reprocessing plants. Numerical experiments were carried out to examine the method above mentioned the estimated subcriticalities agreed well with the true values. The ``Pattern matchingʹʹ method can also be applied to the fuel concentration estimation problem. The estimated fuel distribu- tion agreed reasonably with the true fuel distribution in all cases. It is concluded that the ``Pattern matchingʹʹ method is available critical safety monitor for non-uniform fuel systems in reprocessing plants