A Trajectory Planning Approach via MPC-VOCBF

Model Predictive Control (MPC) has limitations in trajectory planning and obstacle avoidance due to its real-time optimization, such as poor adaptability to dynamic environments. To tackle the aforementioned challenges, we incorporate VOCBF within the MPC framework, thereby shifting safety constraints from reactive distance measures in the spatial domain to proactive dynamic restrictions in the velocity domain. It constructs an improved Control Barrier Function (VOCBF) based on Velocity Obstacle (VO) in the velocity space. By transforming safety constraints from position to velocity space, the robots' ability of dynamic collision avoidance is enhanced. Simulation results show that MPC-VOCBF can guide robots more effectively to reach their target positions than MPC-DC and MPC-CBF methods, greatly improving obstacle avoidance performance in complex environments.