Publications

BibTex format

@article{Phillips:2021:10.1002/nme.6681,
author = {Phillips, TRF and Heaney, CE and Smith, PN and Pain, CC},
doi = {10.1002/nme.6681},
journal = {International Journal for Numerical Methods in Engineering},
pages = {3780--3811},
title = {An autoencoderbased reducedorder model for eigenvalue problems with application to neutron diffusion},
url = {http://dx.doi.org/10.1002/nme.6681},
volume = {122},
year = {2021}
}

Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - Using an autoencoder for dimensionality reduction, this article presents a novel projectionbased reducedorder model for eigenvalue problems. Reducedorder modeling relies on finding suitable basis functions which define a lowdimensional space in which a highdimensional system is approximated. Proper orthogonal decomposition (POD) and singular value decomposition (SVD) are often used for this purpose and yield an optimal linear subspace. Autoencoders provide a nonlinear alternative to POD/SVD, that may capture, more efficiently, features or patterns in the highfidelity model results. Reducedorder models based on an autoencoder and a novel hybrid SVDautoencoder are developed. These methods are compared with the standard PODGalerkin approach and are applied to two test cases taken from the field of nuclear reactor physics.
AU  - Phillips,TRF
AU  - Heaney,CE
AU  - Smith,PN
AU  - Pain,CC
DO  - 10.1002/nme.6681
EP  - 3811
PY  - 2021///
SN  - 0029-5981
SP  - 3780
TI  - An autoencoderbased reducedorder model for eigenvalue problems with application to neutron diffusion
T2  - International Journal for Numerical Methods in Engineering
UR  - http://dx.doi.org/10.1002/nme.6681
UR  - https://onlinelibrary.wiley.com/doi/10.1002/nme.6681
UR  - http://hdl.handle.net/10044/1/88605
VL  - 122
ER  -

Download

DrClaireHeaney

Contact

Location

Summary

Citation

BibTex format

RIS format (EndNote, RefMan)