Publication: Prediction of a reactivity margin for partial refueling of nuclear reactor using artificial neural networks
Дата
2020
Авторы
Journal Title
Journal ISSN
Volume Title
Издатель
Аннотация
© 2020 The Authors. Published by Elsevier B.V.For some types of nuclear reactors (especially research ones) partial refueling is a routine operation when fuel rods are reloading during reactor operation to improve its physical characteristics. Reactivity margin or effective neutron multiplication factor are crucial parameters there because they determine nuclear safety of a facility. Thus for a safety reason a value of effective neutron multiplication factor should be calculated before refueling. A common method consists in neutron physical calculations-simple, but with rather high error. From the other side, precise computer modelling based on Monte-Carlo approach can be used, but this way is very time consuming. In this paper the new approach proposed when artificial neural network used to predict a value of effective neutron multiplication factor using information about fuel burnup in the reactor core as a input data. Training dataset is provided through Monte-Carlo modelling. Optimal layout and parameters selection is considered as well. Results obtained are very perspective for using this approach in real practice at nuclear facilities.
Описание
Ключевые слова
Цитирование
Nakhabov, A. Prediction of a reactivity margin for partial refueling of nuclear reactor using artificial neural networks / Nakhabov, A., Kolesov, V., Soglaev, P. // Procedia Computer Science. - 2020. - 169. - P. 310-313. - 10.1016/j.procs.2020.02.188