Although the annual cane pruning of grapevine yards is a time-consuming job and leads to high labour costs, so far there is no commercial mechanized way to do it. Hence this thesis aims to mechanically design a robotic system that is capable of grapevine cane pruning. This system consists of two similar manipulators and end-effectors, a platform, a computer, a PTO-generator and an agricultural tractor. The platform straddles the vine lines and is fastened to an agricultural tractor three-point hitch. The manipulators are attached to either side of the platform therefore each side of the line is pruned by one of the manipulators. The end-effector is comprised of a cutting and a disinfectant mechanism. After making the cut by the former, the latter mechanism applies some disinfectant liquid to the cut cross section to avoid vine diseases. The energy needs of this robotic system are supplied by a power take-off generator. The manipulator is designed kinematically. In the first stage, the prototype is set to be an articulated redundant 7-DOF manipulator. Then its link lengths are determined through optimizing the volume and isotropy of the robot’s global translational manipulability ellipsoid via Non-dominated Sorting Genetic Algorithm (NSGA-II). The measure of manipulability and the measure of isotropy are used as indices for the volume and the isotropy of the manipulability ellipsoid, respectively. According to previous researches, concerning the aforementioned kinematic indices, the robot spherical wrist is chosen to be Roll-Pitch-Roll type. The final designed robot is named as Prubot. In the previous researches UR5 manipulator is used to prune vines and trees. Thus finally Prubot and UR5 were simulated and eva‎luated through eight sampling-based path planning algorithms to prune a grapevine model. Results demonstrate that Prubot’s performance is generally suitable, and is much better and more successful than UR5 performance, indicating Prubot’s appropriate design.

مصطفي غيور

مهديه غلامي

محمد دانش، حسن موسوي