Learning Trajectories of Hamiltonian Systems with Neural Networks

Loading...
Thumbnail Image

Access rights

openAccess
acceptedVersion

URL

Journal Title

Journal ISSN

Volume Title

A4 Artikkeli konferenssijulkaisussa

Major/Subject

Mcode

Degree programme

Language

en

Pages

12

Series

Artificial Neural Networks and Machine Learning – ICANN 2022, pp. 562–573, Lecture Notes in Computer Science ; Volume 13529

Abstract

Modeling of conservative systems with neural networks is an area of active research. A popular approach is to use Hamiltonian neural networks (HNNs) which rely on the assumptions that a conservative system is described with Hamilton's equations of motion. Many recent works focus on improving the integration schemes used when training HNNs. In this work, we propose to enhance HNNs with an estimation of a continuous-time trajectory of the modeled system using an additional neural network, called a deep hidden physics model in the literature. We demonstrate that the proposed integration scheme works well for HNNs, especially with low sampling rates, noisy and irregular observations.

Description

Other note

Citation

Haitsiukevich, K & Ilin, A 2022, Learning Trajectories of Hamiltonian Systems with Neural Networks. in E Pimenidis, M Aydin, P Angelov, C Jayne & A Papaleonidas (eds), Artificial Neural Networks and Machine Learning – ICANN 2022. Lecture Notes in Computer Science, vol. 13529, Springer, Bristol, UK, pp. 562–573, International Conference on Artificial Neural Networks, Bristol, United Kingdom, 06/09/2022. https://doi.org/10.1007/978-3-031-15919-0_47