Potential-informed path planning with control barrier functions

Path planning is crucial for mobile robots to complete tasks in complex and dynamic environments. Sampling-based methods have gained significant attention in path planning research due to their ability to rapidly find feasible solutions in a given environment. However, these methods often suffer from high sample complexity and struggle to handle unknown obstacles, which limits their real-time applicability. To enhance planning efficiency in dynamic scenes, this study introduces an improved RRT* algorithm, IPIF-RRT*-CBF, which integrates sampling-based planning with control barrier functions. In the sampling phase, the method employs an enhanced potential function-based information sampling strategy to sample and adjust random nodes on the map. During the tree expansion process, the path length and yaw angle are optimized through the active creation of nodes. In the path tracking phase, higher-order CBF constraints are applied within the robot’s field of view to avoid collisions with unknown obstacles and ensure safety. Through simulations and comparisons with current advanced algorithms, the proposed algorithm demonstrates advantages in planning time and safety and can handle unknown obstacles in real time.