Physics-Aware Sparse Learning and Selective Online Adaptation for Euler-Lagrange Robot Dynamics

2026-06-08

AI Summary

This work addresses the limited accuracy of conventional Euler–Lagrange dynamics models under varying payloads, unmodeled couplings, and environmental disturbances, a problem exacerbated by existing learning-based correction methods that often compromise physical structure. To overcome this, the authors propose a structure-preserving residual learning framework that decomposes model mismatch into three physically interpretable components—inertial correction, induced Coriolis terms, and generalized force residuals—each constrained by domain-specific physics and modeled with sparse dependence on historical data. A latent-variable interaction selection mechanism combined with Bayesian linear regression enables selective online adaptation. Validated across mobile, aerial, and robotic manipulator platforms, the approach significantly enhances dynamic prediction accuracy and trajectory tracking performance in time-varying and strongly coupled scenarios while preserving both physical consistency and adaptability.

Abstract

Accurate dynamics models are essential for model-based robotic control, yet nominal Euler--Lagrange models often become inaccurate in the presence of payload variation, unmodeled coupling, friction, aerodynamic effects, and changing operating conditions. Most learning-based correction methods improve prediction accuracy by introducing a single additive residual, but do not preserve the internal mechanical structure of Euler--Lagrange systems. This leads to models that do not preserve symmetry, positive-definiteness, or the coupling between inertia and velocity-dependent terms, which can result in physically inconsistent predictions and reduced reliability when embedded in model-based controllers. We propose a structure-preserving residual learning framework that decomposes model mismatch into an inertia correction, the corresponding induced Coriolis term, and a generalized-force residual. The mechanical component is learned under physical constraints, while the disturbance-sensitive component is represented through a sparse history-dependent latent interaction model and adapted online using Bayesian linear regression. This separation preserves key mechanical structure while restricting adaptation to the part of the dynamics most affected by changing conditions. Experiments across multiple robotic platforms, including mobile, aerial, and manipulator systems, show that the proposed method improves dynamics prediction and trajectory tracking under coupled and time-varying dynamics. These results highlight the value of combining structured residual modeling, compact latent interaction selection, and selective online adaptation for real-world model-based control.

Problem

• Euler-Lagrange dynamics
• structure-preserving learning
• robotic control
• model mismatch
• physical consistency

Innovation

• structure-preserving learning
• Euler-Lagrange dynamics
• sparse latent interaction
• selective online adaptation
• physics-aware modeling