Estimation of Neutronic Performance of a High Power Density Hybrid Reactor by Multilayer Perceptron Neural Networks

Abstract

Artificial neural networks (ANNs) have recently been utilized in the nuclear technology applications since they are fast, precise and flexible vehicles to modeling, simulation and optimization. This paper presents a new approach based on multilayer perceptron neural networks (MLPNNs) for the estimation of some important neutronic parameters (net Pu-239 production, tritium breeding ratio, cumulative fissile fuel enrichment, and fission rate) of a high power density fusion-fission (hybrid) reactor using light water reactor (LWR) spent fuel. A comparison of the results obtained by the MLPNNs and those found by using the code (Scale 4.3) was carried out. The results pointed out that the MLPNNs trained with least mean squares (LMS) algorithm could provide an accurate computation of the main neutronic parameters for the high power density reactor.