Neutron spectrum unfolding using two architectures of convolutional neural networks

We deploy artificial neural networks to unfold neutron spectra from measured energy-integrated quantities. These neutron spectra represent an important parameter allowing to compute the absorbed dose and the kerma to serve radiation protection in addition to nuclear safety. The built architectures are inspired from convolutional neural networks. The first architecture is made up of residual transposed convolution's blocks while the second is a modified version of the U-net architecture. A large and balanced dataset is simulated following realistic physical constraints to train the architectures in an efficient way. Results show a high accuracy prediction of neutron spectra ranging from thermal up to fast spectrum. The dataset processing, the attention paid to performances' metrics and the hyperoptimization are behind the architectures' robustness. (c) 2023 Korean Nuclear Society, Published by Elsevier Korea LLC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Automated summary

In this study, artificial neural networks are used to unfold neutron spectra from measured energy-integrated quantities. These spectra are important for calculating absorbed dose and kerma for radiation protection and nuclear safety. The architectures used are inspired by convolutional neural networks, with one architecture using residual transposed convolution blocks and the other being a modified version of the U-net architecture. A large and balanced simulated dataset is used to train the architectures effectively, considering realistic physical constraints. The results demonstrate high accuracy in predicting neutron spectra from thermal to fast spectrum, with dataset processing, attention to performance metrics, and hyperoptimization contributing to the robustness of the architectures.