Abstract:Through research, it was found that octopus has multiple tentacles and more than 200 suction cups on each tentacle, which can effectively avoid the prey from breaking free by the combination of tentacle grasping and suction cup adsorption without knowing the shape of the prey. In order to pick the end of the robot to meet the different shapes, sizes and growth conditions of the apple picking work, the octopus prey way was drawn on, according to the general characteristics of the Shaanxi Fuji apple parameters, the picking robot adsorption mechanism, gripping mechanism, power transmission device and other key parts of the design optimization and its size parameters to determine the design of a set of suction cups adsorption and finger gripping as a combination of pneumatic apple picking robot were completed. A composite pneumatic apple picking manipulator integrating suction cup adsorption and finger gripping was designed. According to the three-dimensional configuration of the manipulator, through the kinematics and dynamics analysis of the suction cup adsorption and finger gripping, the kinematics and dynamics simulation models were established and simulation tests were carried out, and it can be seen from the simulation results that the movement trends of the end of the adsorption mechanism and the end of the gripping mechanism had a better consistency, and the relative slippage between the manipulator and apples was within the range of 0.19~2.28mm. The above simulation results showed that the manipulator had good flexibility of movement and stability of adsorption and grasping. The experimental platform of the six-armed apple harvesting robot was utilized to conduct field trials in an orchard, focusing on the picking of apples. The test classified the apples available for harvesting by the robot into four categories: vertically grown and unobstructed, vertically grown with minimal obstruction, non-vertically grown and unobstructed, and non-vertically grown with minimal obstruction. A comprehensive evaluation was performed based on these categories. The results of the experiment demonstrate that the harvesting hand achieved an overall success rate of 84.7% and a damage rate of only 0.88%, showcasing its capability to accomplish apple picking tasks in intricate orchard environments.