Work

Few-group Cross Sections Modeling by Artificial Neural Networks

Year: 2021

Type: conference-paper

Abstract: This work deals with the modeling of homogenized few-group cross sections by Artificial Neural Networks (ANN). A comprehensive sensitivity study on data normalization, network architectures and training hyper-parameters specifically for Deep and Shallow Feed Forward ANN is presented. The optimal models in terms of reduction in the library size and training time are compared to multi-linear interpolation on a Cartesian grid. The use case is provided by the OECD-NEA Burn-up Credit Criticality Benchmark [1]. The Pytorch [2] machine learning framework is used.

Source: EPJ Web of Conferences

DOI: 10.1051/epjconf/202124706029

Authors: E. Szames, K. Ammar, D. Tomatis, J.M. Martinez

Institutions: Commissariat à l'Énergie Atomique et aux Énergies Alternatives, Université Paris-Saclay, Direction des Energies, CEA Paris-Saclay, Service d'Étude des Réacteurs et de Mathématiques Appliquées

Language: English

Topic: Nuclear reactor physics and engineering

Subfield: Nuclear Energy and Engineering

Field: Energy

Domain: Physical Sciences

Open Access status: diamond

OpenAlex ID: W3131111245

FWCI: 8.073

Cites: 13

Cited by: 12

Related to: 10