Beyond Terminal Constraints: Reliable Robot Control with Learned Safe Sets
Beyond Terminal Constraints: Reliable Robot Control with Learned Safe Sets
This research paper introduces new classes of safe sets for robot control that are broader than classical control-invariant sets. The authors demonstrate that safety can be guaranteed by constraining different nodes of the prediction horizon in Model Predictive Control (MPC) schemes, rather than relying solely on terminal state constraints. Their methods ensure recursive feasibility and can employ a safe-abort strategy when the safe set is an inner approximation of a control-invariant set. Extensive simulations on a 6-joint robot manipulator show fewer constraint violations compared to state-of-the-art methods while maintaining comparable tracking performance and computational efficiency. Hardware experiments further support the practicality of these schemes for real-world applications.
The authors propose novel safe set formulations for Model Predictive Control and Reinforcement Learning that relax strict control-invariance requirements, validated via simulations on a 6-joint manipulator and preliminary hardware experiments.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S4
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ds can ensure safety through a safe-abort strategy, even in the absence of recursive feasibility. We validate our approaches in extensive simulations on a 6-joint robot manipulator, showing fewer constraint violations than state-of-the-art methods while maintaining comparable tracking performance and computational efficiency. Hardware experiments further demonstrate the practicality of our schemes for real-world robo
Open source S4
arXiv:2611.06196v1 Announce Type: new Abstract: Ensuring safety in advanced robot control methods such as Model Predictive Control (MPC) and Reinforcement Learning remains an open challenge. A common approach relies on control-invariant sets, yet their exact computation for nonlinear systems is generally intractable, and numerical approximations compromise theoretical guarantees. In this work, we show that safety doe
Open source S2
Source：arXiv Robotics — research abstracts · arxiv.org