We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

physics.comp-ph

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Physics > Computational Physics

Title: Data Reduction in Deterministic Neutron Transport Calculations Using Machine Learning

Abstract: Neutron cross section matrices for fission and scattering data are required for each material, temperature, and enrichment level to calculate the neutron transport equation accurately. This information can be a limiting factor when using the multigroup discrete ordinates (SN) method when the number of energy groups is large. Machine Learning (ML) can be used to replace the need for the cross section matrices by reproducing the function that maps the scalar flux to the scattering and fission sources. Through the use of autoencoders and Deep Jointly-Informed Neural Networks (DJINN), the data storage requirements are reduced by 94% of the original data for a 618 group problem. This is accomplished while preserving the scalar flux, maintaining generality, and decreasing wall clock times.
Subjects: Computational Physics (physics.comp-ph)
Cite as: arXiv:2205.05165 [physics.comp-ph]
  (or arXiv:2205.05165v1 [physics.comp-ph] for this version)

Submission history

From: Benjamin Whewell [view email]
[v1] Tue, 10 May 2022 20:54:52 GMT (3991kb,D)

Link back to: arXiv, form interface, contact.