AI Libraries Meet Neutron Transport in Boltzmann Equation

Authors T. R. F. Phillips et al. present their work on solving the Boltzmann transport equation using AI libraries in the paper Solving the Discretised Boltzmann Transport Equations using Neural Networks: Applications in Neutron Transport.

• The method leverages optimized AI library software for different computer architectures.
• Introduces a space-angle multigrid solution method.
• Develops a Convolutional Finite Element Method (ConvFEM) for high order finite elements.
• Proposes a novel non-linear Petrov-Galerkin method to introduce dissipation anisotropically.

This paper is an excellent example of how AI libraries can be adapted for complex physical processes and could lead to significant advancements in the field of neutron transport and related areas.