Reduced-order Control and Geometric Structure of Learned Lagrangian Latent Dynamics

TL;DR

This paper develops a structure-preserving reduced-order control framework for high-dimensional Lagrangian systems, enabling stable and convergent control with learned dynamics and physical structure, validated through simulations and real-world experiments.

Contribution

It introduces a novel latent control approach that combines structure-preserving model reduction with learned dynamics for high-dimensional systems, providing stability guarantees.

Findings

The proposed controller achieves stable tracking in simulated high-dimensional systems.

Experimental validation confirms the controller's effectiveness on real-world systems.

The framework quantifies modeling errors and offers interpretable stability conditions.

Abstract

Model-based controllers can offer strong guarantees on stability and convergence by relying on physically accurate dynamic models. However, these are rarely available for high-dimensional mechanical systems such as deformable objects or soft robots. While neural architectures can learn to approximate complex dynamics, they are either limited to low-dimensional systems or provide only limited formal control guarantees due to a lack of embedded physical structure. This paper introduces a latent control framework based on learned structure-preserving reduced-order dynamics for high-dimensional Lagrangian systems. We derive a reduced tracking law for fully actuated systems and adopt a Riemannian perspective on projection-based model-order reduction to study the resulting latent and projected closed-loop dynamics. By quantifying the sources of modeling error, we derive interpretable conditions for stability and convergence. We extend the proposed controller and analysis to underactuated systems by introducing learned actuation patterns. Experimental results on simulated and real-world systems validate our theoretical investigation and the accuracy of our controllers.