An autonomous obstacle avoidance and path planning method for fruit-picking UAV in orchard environments

Abstract

In orchard environments, compared with picking robotic arms, improving the efficiency and safety of the fruit-picking unmanned aerial vehicle (UAV) becomes more challenging. In this paper, an autonomous obstacle avoidance and path planning method based on LiDAR data is proposed for the self-built fruit-picking UAV. First, a LiDAR static-dynamic dual map construction scheme is designed. Using the original point cloud data from LiDAR, a time-accumulated local point cloud map is generated to provide orchard obstacle information for path planning. Then, an improved hybrid A* algorithm based on the B-spline curve is proposed. This algorithm not only comprehensively takes into account the impact of surrounding branches on the flight of the picking UAV near the target fruit bunch, but also ensures that the planned path meets the specific action requirements of the picking UAV when picking the target fruit bunch. The experimental results demonstrate that the proposed map construction scheme significantly reduces the computational power requirements and collision detection time. Moreover, the path planning algorithm effectively guides the UAV and its attached picking actuator to successfully navigate around obstacles, enabling efficient picking of the target fruit bunch. Indicating that the proposed method provides a feasible solution for task execution of the fruit-picking UAV in complex orchard environments.