Stable Port-Hamiltonian Neural Networks

Fabian J. Roth, Dominik K. Klein, Maximilian Kannapinn, Jan Peters, Oliver Weeger · 2025

Read Stable Port-Hamiltonian Neural Networks by Fabian J. Roth, Dominik K. Klein, Maximilian Kannapinn, Jan Peters, Oliver Weeger (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-2502-02480.txt.

Description

Fabian Roth and co-authors teach neural networks physics rules so their forecasts stay stable, in a 2025 paper. Networks that learn motion from data alone often drift into impossible behavior. They can also fail when asked to predict beyond the data they saw. The team's design builds in energy saving and energy loss from the start.

The design rests on a classic physics frame that tracks how energy moves through a system. A built-in proof keeps the learned motion stable over time. In tests with sparse data, the model stayed sound and beat plain data-driven rivals. It also handled rich simulation data from coupled physics problems. The authors see it as a safer base for stand-in models.

Catalog record

CREATORFabian J. Roth, Dominik K. Klein, Maximilian Kannapinn, Jan Peters, Oliver Weeger
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE69 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-02480

Browse the full library · About the Tower of Babel