Control Barrier Function
Control Barrier Functions (CBFs) are mathematical tools used to guarantee the safety of dynamical systems by ensuring that trajectories remain within a predefined "safe set," a key objective being to maximize the size of this safe set while respecting input constraints. Current research focuses on improving CBF's robustness to uncertainties (e.g., disturbances, noisy sensor data) and integrating them with various control architectures like Model Predictive Control (MPC) and reinforcement learning, often employing neural networks for adaptive parameter tuning or online CBF synthesis. This work is significant for enabling safe and reliable operation of autonomous systems in complex and unpredictable environments, with applications ranging from robotics and autonomous driving to aerospace.
Papers
Learning responsibility allocations for multi-agent interactions: A differentiable optimization approach with control barrier functions
Isaac Remy, David Fridovich-Keil, Karen Leung
Safe Reinforcement Learning Filter for Multicopter Collision-Free Tracking under disturbances
Qihan Qi, Xinsong Yang, Gang Xia
Disturbance Observer-based Control Barrier Functions with Residual Model Learning for Safe Reinforcement Learning
Dvij Kalaria, Qin Lin, John M. Dolan