Citrus Picking Robot with Collision-free Picking Path Planning

A real-time obstacle avoidance path planning method based on improved rapidly-exploring random tree (RRT) algorithm is proposed for citrus picking robot manipulators in unstructured environments. Based on the specific structure and obstacle characteristics of a self-developed 6-degree-of-freedom citrus picking robot manipulator, kinematic analysis and obstacle modeling of the manipulator were carried out. To address the drawback of RRT algorithm and to cope with the fact that the RRTConnect algorithm requires repeated and excessive searching in complex environments, which leads to nodes being too close to the edge crossing, a fast path planning algorithm for the manipulator based on sparse dead point preservation strategy was proposed. Several algorithms were compared and validated in a two-dimensional and three-dimensional environment. The results show that the improved RRTConnect algorithm greatly reduces the number of collision detections, improves planning speed, and proves the excellent performance of the improved algorithm in complex environments. This method is applied to the real-time obstacle avoidance task of a harvesting robot manipulator with known end position, obstacle position, and target position. Simulation and experimental research results show that this method is simple, has good real-time performance, can effectively avoid obstacles, and successfully reach the target position. It is suitable for automatic citrus picking in natural growth conditions.