Development and performance evaluation of orchard tethered flexible pear fruit picking UAV system

With the growing shortage of rural labor, orchard mechanized picking has become an inevitable trend. However, existing domestic and international picking robots are mostly based on ground mobile platforms, only suitable for low-height fruit picking. High-position fruits on tall trees still rely on manual picking, which is difficult and high-risk. To address this issue, this study proposes a tethered UAV-based picking method and conducts research on aerial picking of "Fojianxi" pears in a standardized orchard. First, the design requirements and working principles of the picking platform were determined via analyzing the standardized pear orchard environment; design criteria for a lightweight end effector and the optimal picking method were identified based on pear physical properties and comparative analysis of picking technologies. Second, structural design was completed for the quadrotor UAV frame, end effector, and connecting rods, with structural reliability verified by ANSYS. Component selection for the flight platform was performed, and a tethered power supply system was constructed. Gravity balancing using onboard batteries was implemented to minimize the impact of end effector-induced eccentric loads on UAV stability. A mathematical model of the picking UAV was established, and a targeted PID controller for the eccentric-center-of-gravity quadrotor was designed, with its effectiveness validated by MATLAB simulations. Finally, simulated picking experiments were conducted in a standardized pear orchard. Results showed that the initial tethered UAV picking system achieved a 59.26% success rate; the optimized system exhibited higher stability and better fruit clamping performance, with the success rate increased to 75%, enabling effective picking of fruits at various heights.