Structure-preserving neural networks - Archive ouverte HAL Access content directly
Journal Articles Journal of Computational Physics Year : 2020

Structure-preserving neural networks

Abstract

We develop a method to learn physical systems from data that employs feedforward neural networks and whose predictions comply with the first and second principles of thermodynamics. The method employs a minimum amount of data by enforcing the metriplectic structure of dissipative Hamiltonian systems in the form of the so-called General Equation for the Non-Equilibrium Reversible-Irreversible Coupling, GENERIC (Öttinger and Grmela (1997) [36]). The method does not need to enforce any kind of balance equation, and thus no previous knowledge on the nature of the system is needed. Conservation of energy and dissipation of entropy in the prediction of previously unseen situations arise as a natural by-product of the structure of the method. Examples of the performance of the method are shown that comprise conservative as well as dissipative systems, discrete as well as continuous ones.
Fichier principal
Vignette du fichier
PIMM_JCP_ 2020_CHINESTA.pdf (541.33 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03019811 , version 1 (23-11-2020)

Identifiers

Cite

Quercus Hernández, Alberto Badías, David González, Francisco Chinesta, Elías Cueto. Structure-preserving neural networks. Journal of Computational Physics, 2020, pp.1-16. ⟨10.1016/j.jcp.2020.109950⟩. ⟨hal-03019811⟩
72 View
84 Download

Altmetric

Share

Gmail Facebook X LinkedIn More