Enabling scientific machine learning in MOOSE using Libtorch

German, Peter; Yushu, Dewen

doi:10.1016/j.softx.2023.101489

Enabling scientific machine learning in MOOSE using Libtorch

Journal Article · Mon Aug 14 00:00:00 EDT 2023 · SoftwareX

DOI:https://doi.org/10.1016/j.softx.2023.101489· OSTI ID:2331444

^[1]; ^[1]

Idaho National Laboratory (INL), Idaho Falls, ID (United States)

A neural-network-based machine learning interface has been developed for the Multiphysics Object-Oriented Simulation Environment (MOOSE). The interface relies on Libtorch, the C++ front-end of PyTorch, and enables an online interaction between modern machine learning algorithms and all the existing simulation, modeling, and analysis processes available in MOOSE. New capabilities in MOOSE include the native generation and training of artificial neural networks together with options to load pretrained neural networks in TorchScript format. Furthermore, the MOOSE stochastic tools module (MOOSE-STM) has been enhanced with neural network-based surrogate and reduced-order model generation options for efficient stochastic analyses. Lastly, a reinforcement learning capability has been added to MOOSE-STM for the interactive control and optimization of complex multiphysics problems.

Research Organization:: Idaho National Laboratory (INL), Idaho Falls, ID (United States)

Sponsoring Organization:: USDOE Laboratory Directed Research and Development (LDRD) Program

Grant/Contract Number:: AC07-05ID14517

OSTI ID:: 2331444

Report Number(s):: INL/JOU--22-69491-Revision-0

Journal Information:: SoftwareX, Journal Name: SoftwareX Journal Issue: - Vol. 23; ISSN 2352-7110

Publisher:: ElsevierCopyright Statement

Country of Publication:: United States

Language:: English

References (8)

libMesh : a C++ library for parallel adaptive mesh refinement/coarsening simulations Kirk, Benjamin S.; Peterson, John W.; Stogner, Roy H. Engineering with Computers, Vol. 22, Issue 3-4 https://doi.org/10.1007/s00366-006-0049-3	journal	November 2006
Physics-based multiscale coupling for full core nuclear reactor simulation Gaston, Derek R.; Permann, Cody J.; Peterson, John W. Annals of Nuclear Energy, Vol. 84 https://doi.org/10.1016/j.anucene.2014.09.060	journal	October 2015
Rapid multiphase-field model development using a modular free energy based approach with automatic differentiation in MOOSE/MARMOT Schwen, D.; Aagesen, L. K.; Peterson, J. W. Computational Materials Science, Vol. 132 https://doi.org/10.1016/j.commatsci.2017.02.017	journal	May 2017
EventGraD: Event-triggered communication in parallel machine learning Ghosh, Soumyadip; Aquino, Bernardo; Gupta, Vijay Neurocomputing, Vol. 483 https://doi.org/10.1016/j.neucom.2021.08.143	journal	April 2022
Validating the BISON fuel performance code to integral LWR experiments Williamson, R. L.; Gamble, K. A.; Perez, D. M. Nuclear Engineering and Design, Vol. 301 https://doi.org/10.1016/j.nucengdes.2016.02.020	journal	May 2016
2.0 - MOOSE: Enabling massively parallel multiphysics simulation Lindsay, Alexander D.; Gaston, Derek R.; Permann, Cody J. SoftwareX, Vol. 20 https://doi.org/10.1016/j.softx.2022.101202	journal	December 2022
Learning representations by back-propagating errors Rumelhart, David E.; Hinton, Geoffrey E.; Williams, Ronald J. Nature, Vol. 323, Issue 6088 https://doi.org/10.1038/323533a0	journal	October 1986
Rattlesnake: A MOOSE-Based Multiphysics Multischeme Radiation Transport Application Wang, Yaqi; Schunert, Sebastian; Ortensi, Javier Nuclear Technology, Vol. 207, Issue 7 https://doi.org/10.1080/00295450.2020.1843348	journal	April 2021

Similar Records

Scientific Machine Learning using MOOSE

Conference · Wed Jul 26 20:00:00 EDT 2023 · OSTI ID:2305504

MOOSE ProbML: Parallelizable Probabilistic Machine Learning and Uncertainty Quantification Capabilities

Conference · Mon Mar 10 20:00:00 EDT 2025 · OSTI ID:3028337

MOOSE ProbML: Parallelized probabilistic machine learning and uncertainty quantification for computational energy applications

Journal Article · Sun Dec 21 19:00:00 EST 2025 · Journal of Computational Science · OSTI ID:3012815

Related Subjects

97 MATHEMATICS AND COMPUTING
MOOSE
libtorch
reinforcement learning
scientific machine learning

Enabling scientific machine learning in MOOSE using Libtorch

Citation Formats

References (8)

Similar Records

Related Subjects