BubbleCBF: Constructive Safety for Second-Order Systems and Application to Autonomous Robot Navigation in Unknown Environments
We develop a constructive approach for designing feasible safety constraints for second-order control-affine systems.
ProofPaper ↗
Key points
- Satisfying safety constraints for systems with second-order dynamics is challenging because a position constraint may require infeasible deceleration.
- Starting from a signed distance function (SDF), we construct collision-free spheres, i.e., bubbles, in position space, and use their quadratic geometry and overlap to construct a velocity-dependent safety constraint.
- Under mild assumptions on the dynamics, bubble overlap, and design parameters, we prove that the quadratic-program-based controller is well defined, continuous in the system state, and renders the set forward invariant.
- We demonstrate that our BubbleCBF, applied to a differential-drive robot, achieves safe autonomous navigation in unknown environments.
Sources (1)
- [1]BubbleCBF: Constructive Safety for Second-Order Systems and Application to Autonomous Robot Navigation in Unknown EnvironmentsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 14, 08:24 PM
We develop a constructive approach for designing feasible safety constraints for second-order control-affine systems.
Satisfying safety constraints for systems with second-order dynamics is challenging because a position constraint may require infeasible deceleration.
Extractive summary: sentences quoted from the sources.