Non-intrusive reduced-order modeling for dynamical systems with spatially localized features

Leonidas Gkimisis, Nicole Aretz, Marco Tezzele, Thomas Richter, Peter Benner, Karen E. Willcox · 2025

Read Non-intrusive reduced-order modeling for dynamical systems with spatially localized features by Leonidas Gkimisis, Nicole Aretz, Marco Tezzele, Thomas Richter, Peter Benner, Karen E. Willcox (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-04400.txt.

Description

Six authors, led by Leonidas Gkimisis, present a 2025 paper on faster models for systems that change over time and space. Full simulations of such systems can be slow, so engineers build reduced models that mimic them at lower cost. Some systems have sharp local features, which make their data decay slowly and resist simple reduction.

The team's framework learns the model from data alone, without stepping inside the original code. It splits the domain into two parts. One part decays fast and is learned with Operator Inference, a lean method. The other part holds the sharp features and is learned with a sparse full-order method that keeps fine detail. Joining the two gives fast forecasts for new conditions beyond the training samples. The approach pairs low cost in smooth zones with high detail where it counts.

Catalog record

CREATORLeonidas Gkimisis, Nicole Aretz, Marco Tezzele, Thomas Richter, Peter Benner, Karen E. Willcox
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE80 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-04400

Browse the full library · About the Tower of Babel