Turning Circle-Based Control Barrier Function for Efficient Collision Avoidance of Nonholonomic Vehicles

Generating safe and efficient collision-free trajectories and control inputs is essential for autonomous vehicles. While control barrier functions (CBFs) are widely used to ensure safety, conventional CBFs based on the Euclidean distance often neglect the nonholonomic constraints commonly found in autonomous vehicles, leading to inefficient or overly conservative maneuvers. This brief proposes a turning circle-based CBF (TC-CBF) that explicitly accounts for the vehicle’s heading and turning limitations by considering the proximity between its turning circles and surrounding obstacles. By integrating the TC-CBF with model predictive control (MPC), we develop the MPC-TCCBF framework, which enables the efficient trajectory planning and the computation of control inputs. Simulations and hardware experiments using unicycle models and underactuated surface vehicles demonstrate that MPC-TCCBF enhances the collision avoidance performance.