pipette
ESEspañol

Port-Hamiltonian Neural Networks for Systems with Multiple Asymptotically Stable Equilibria

Simon Heilig, Jens P\"uttschneider, Mohammad Itani, Asja Fischer, Timm Faulwasser

Preprint

In the authors' words

Stable port-Hamiltonian neural networks certify asymptotic stability by construction. Yet, their Hamiltonian is a global Lyapunov function with a single global minimum, so they can represent only dynamic systems with {one} attractor. We demonstrate that this excludes even simple systems with energy landscapes forming a double well, and we overcome the restriction by parametrising the Hamiltonian as a {product} of Bregman divergences generated by one input-convex network. We prove that the resulting model is locally Lyapunov stable, that the coexistence of stable equilibria forces additional non-asymptotically-stable equilibria to exist, that all equilibria lie in a bounded region, and under a hyperbolicity assumption that almost-everywhere stability holds. On three systems our approach is able to recover the energy surface characteristics and improve the convergence speed by 1.8-8.5.

Main resultLimitation the authors admit

Appeared: Friday, October 2. arXiv. Preprint, not yet peer-reviewed.

Authors' comment: Accepted at NeurIPS 2026 Workshop: AXIOM - Foundations of Efficient Deep Learning