Co-Learning Port-Hamiltonian Systems and Optimal Energy-Shaping Control

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kamboj, Ankur, Dey, Biswadip, Srivastava, Vaibhav
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915983299444736
author Kamboj, Ankur
Dey, Biswadip
Srivastava, Vaibhav
author_facet Kamboj, Ankur
Dey, Biswadip
Srivastava, Vaibhav
contents We develop a physics-informed learning framework for energy-shaping control of port-Hamiltonian (pH) systems from trajectory data. The proposed approach co-learns a pH system model and an optimal energy-balancing passivity-based controller (EB-PBC) through alternating optimization with policy-aware data collection. At each iteration, the system model is refined using trajectory data collected under the current control policy, and the controller is re-optimized on the updated model. Both components are parameterized by neural networks that embed the pH dynamics and EB-PBC structure, ensuring interpretability in terms of energy interactions. The learned controller renders the closed-loop system inherently passive and provably stable, and exploits passive plant dynamics without canceling the natural potential. A dissipation regularization enforces strict energy decay during training, thereby enhancing robustness to sim-to-real gaps. The proposed framework is validated on state-regulation and swing-up tasks for planar and torsional pendulum systems.
format Preprint
id arxiv_https___arxiv_org_abs_2604_26172
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Co-Learning Port-Hamiltonian Systems and Optimal Energy-Shaping Control
Kamboj, Ankur
Dey, Biswadip
Srivastava, Vaibhav
Systems and Control
Artificial Intelligence
Machine Learning
Optimization and Control
We develop a physics-informed learning framework for energy-shaping control of port-Hamiltonian (pH) systems from trajectory data. The proposed approach co-learns a pH system model and an optimal energy-balancing passivity-based controller (EB-PBC) through alternating optimization with policy-aware data collection. At each iteration, the system model is refined using trajectory data collected under the current control policy, and the controller is re-optimized on the updated model. Both components are parameterized by neural networks that embed the pH dynamics and EB-PBC structure, ensuring interpretability in terms of energy interactions. The learned controller renders the closed-loop system inherently passive and provably stable, and exploits passive plant dynamics without canceling the natural potential. A dissipation regularization enforces strict energy decay during training, thereby enhancing robustness to sim-to-real gaps. The proposed framework is validated on state-regulation and swing-up tasks for planar and torsional pendulum systems.
title Co-Learning Port-Hamiltonian Systems and Optimal Energy-Shaping Control
topic Systems and Control
Artificial Intelligence
Machine Learning
Optimization and Control
url https://arxiv.org/abs/2604.26172