An Improved RRT Algorithm Based on Bézier Curves for Logistics Delivery UAV Path Planning

This paper investigates the path-planning problem of unmanned aerial vehicles (UAVs) for logistics delivery in urban environments. The impact of real-time obstacle avoidance and path smoothness on the flyability of UAVs remains a challenge in existing research. To address the issue that the path generated by the traditional Rapidly exploring Random Tree (RRT) algorithm exhibits a sudden slope change at connection points, which makes the UAV non-flyable, this paper proposes an improved algorithm that combines the traditional RRT algorithm with Bézier curves. The proposed real-time path generation strategy consists of two stages. The first stage constructs the environment model. The second stage integrates the RRT algorithm with Bézier curves, enabling the generated route to achieve real-time obstacle avoidance while being smooth and free of curvature discontinuities. Simulation experiments compare the improved algorithm with the traditional RRT algorithm and global path optimization methods. The experimental results show that the improved algorithm has the advantage of real-time obstacle avoidance, and the generated route is smooth at connection points with no curvature discontinuities, thereby ensuring good flyability.