analiza aiqbench
This research advances robotic control by integrating reinforcement learning with disturbance observers to handle parameter uncertainty and friction in manipulators. By enforcing input-to-state stability bounds, the framework ensures provable safety while achieving significant tracking error reductions on real hardware during zero-shot sim-to-real transfer. The approach effectively isolates model residuals for targeted learning, enabling robust adaptation to unseen disturbances like base vibrations without retraining.
Stability-aware Residual Reinforcement Learning Framework for Robotic Manipulator Disturbance Compensation
arXiv:2609.21307v1 Announce Type: new Abstract: Although conventional controllers and disturbance observers (DOBs) are the standard for precision tracking in manipulators, they suffer from parameter uncertainty, nonlinear friction, and compound disturbances. This study proposes a residual reinforcement learning DOB framework that pairs an analytical observer with an RL policy. The deterministic baseline operates within a reliable region, whereas the RL policy explicitly targets the residuals that the model cannot capture. To make this compensation disturbance-aware, an estimator network aligns the observation history with a privileged disturbance context, organizing the latent space by disturbance regime and enabling rapid adaptation across disturbance transitions. To guarantee stability, we derived and enforced a state-dependent action bound on the RL policy from an input-to-state stability (ISS) analysis such that the closed loop provably confines the tracking error to a certified envelope for arbitrary policy outputs. Experiments on a 6-DOF manipulator demonstrated consistent improvements in disturbance estimation and tracking, including a 27.8% tracking-error reduction on real hardware under zero-shot sim-to-real transfer and a 38.0% reduction under a base-vibration disturbance that was not observed during training.
Abstract:Although conventional controllers and disturbance observers (DOBs) are the standard for precision tracking in manipulators, they suffer from parameter uncertainty, nonlinear friction, and compound disturbances. This study proposes a residual reinforcement learning DOB framework that pairs an analytical observer with an RL policy. The deterministic baseline operates within a reliable region, whereas the RL policy explicitly targets the residuals that the model cannot capture. To make this compensation disturbance-aware, an estimator network aligns the observation history with a privileged disturbance context, organizing the latent space by disturbance regime and enabling rapid adaptation across disturbance transitions. To guarantee stability, we derived and enforced a state-dependent action bound on the RL policy from an input-to-state stability (ISS) analysis such that the closed loop provably confines the tracking error to a certified envelope for arbitrary policy outputs. Experiments on a 6-DOF manipulator demonstrated consistent improvements in disturbance estimation and tracking, including a 27.8% tracking-error reduction on real hardware under zero-shot sim-to-real transfer and a 38.0% reduction under a base-vibration disturbance that was not observed during training.
Robotics (cs.RO)
arXiv:2609.21307 [cs.RO]
(or arXiv:2609.21307v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.21307 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 266 cuvinte
Subiecte conexe
Res-HIL: Human-Guided Residual Reinforcement Learning for Sample-Efficient Dexterous Manipulation
arXiv · 25 sept. 2026
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · 17 sept. 2026
Safety-aware Skill Adaptation for Reinforcement Learning in Dynamic Environments
arXiv · 11 sept. 2026
TORL-VLA: Tactile Guided Online Reinforcement Learning for Contact-Rich Manipulation
arXiv · 11 iun. 2026