Learn Structure, Adapt on the Fly: Multi-Scale Residual Learning and Online Adaptation for Aerial Manipulators

2026-03-12

AI Summary

This work addresses the challenge of accurately modeling the nonstationary, multiscale residual dynamics exhibited by autonomous aerial manipulators under rapid configuration changes and abrupt payload variations—phenomena poorly captured by conventional models. To this end, the authors propose a prediction-adaptation framework that integrates a Factorized Dynamics Transformer (FDT) to explicitly decouple short-term inertial effects from long-term aerodynamic influences, alongside a latent-space Linear Residual Adapter (LRA) based on recursive least squares for rapid online adaptation. The proposed approach significantly enhances residual dynamics modeling accuracy, disturbance rejection speed, and closed-loop tracking performance while strictly preserving real-time computational constraints, outperforming existing state-of-the-art methods.

Technology Category

• Intelligent Robots: State Estimation
• Multiagent Systems: Adversarial Agents
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Autonomous Aerial Manipulators (AAMs) are inherently coupled, nonlinear systems that exhibit nonstationary and multiscale residual dynamics, particularly during manipulator reconfiguration and abrupt payload variations. Conventional analytical dynamic models rely on fixed parametric structures, while static data-driven model assume stationary dynamics and degrade under configuration changes and payload variations. Moreover, existing learning architectures do not explicitly factorize cross-variable coupling and multi-scale temporal effects, conflating instantaneous inertial dynamics with long-horizon regime evolution. We propose a predictive-adaptive framework for real-time residual modeling and compensation in AAMs. The core of this framework is the Factorized Dynamics Transformer (FDT), which treats physical variables as independent tokens. This design enables explicit cross-variable attention while structurally separating short-horizon inertial dependencies from long-horizon aerodynamic effects. To address deployment-time distribution shifts, a Latent Residual Adapter (LRA) performs rapid linear adaptation in the latent space via Recursive Least Squares, preserving the offline nonlinear representation without prohibitive computational overhead. The adapted residual forecast is directly integrated into a residual-compensated adaptive controller. Real-world experiments on an aerial manipulator subjected to unseen payloads demonstrate higher prediction fidelity, accelerated disturbance attenuation, and superior closed-loop tracking precision compared to state-of-the-art learning baselines, all while maintaining strict real-time feasibility.

Problem

• Aerial Manipulators
• Nonstationary Dynamics
• Multi-scale Residual Dynamics
• Cross-variable Coupling
• Payload Variations

Innovation

• Factorized Dynamics Transformer
• Online Adaptation
• Multi-Scale Residual Learning
• Latent Residual Adapter
• Aerial Manipulators