Research On The Precise Picking Robot Of High-quality Tea Sprout With Multi-arm Coordination

With the improvement of people’s living standards,the demand for high-end high-quality tea is increasing,but the difficulty of picking has seriously restricted its development.High-quality tea picking standards are strict,inefficient,and there is a severe shortage of labor,and workers are generally aging.Most of the mechanized tea picking adopts a "one size fits all" type of harvesting,which is only suitable for bulk teas with low added value,and cannot replace manual labor in the picking of high-quality teas.Therefore,precise and high-speed mechanized picking of tender buds has become a key issue restricting the development of high-quality tea.Focusing on the accurate and high-speed picking of high-quality tea buds,a crawler-type picking robot adapted to the tea garden environment was designed,a multi-manipulator collaborative picking scheme was proposed,and an efficient picking operation method was studied.Design a picking robot suitable for tea gardens,establish a crawler motion control model,and achieve precise tracking.According to the tea garden planting specifications,combined with the crawler mechanism,the main structure of the picking robot with strong adaptability and easy production is designed.Analyze the functional requirements of precise picking,and design six subsystems of robot main controller,drive control,image acquisition,sprout picking,sensing assistance,and human-computer interaction,so that it can stably complete tasks such as movement,precise picking,and perception interaction.The track drive algorithm based on FOC(vector control)is used for motion control.The simulation shows that the response speed of the motor is less than 20 ms,and the torque control error is less than 1%.The precise control of the trajectory is realized,which provides a platform foundation for subsequent research on picking.Based on the synovial membrane control algorithm,a trajectory tracking model of the picking arm is constructed to realize the precise picking of single tea buds.Determine the DH parameter matrix of the picking manipulator,establish the kinematic equation,and inversely solve it to achieve precise control of single-point picking.In response to the requirement of continuous picking at multiple picking points,the synovial control algorithm is applied,the seven-segment S-curve is used to constrain the trajectory of the robotic arm,and the trajectory tracking control is carried out,which realizes the smooth trajectory of the picking process and avoids violent collision with the tea crown.,and the control error is controlled within 2 mm,which verifies the correctness and feasibility of the multi-point continuous picking scheme.In order to further improve the picking speed,a multi-agent reinforcement learning algorithm is used to build a multi-manipulator collaborative picking model.According to the specifications of the manipulators and the distribution status of the sprouts,a rectangular layout scheme of four picking manipulators is designed,and a picking point division model is constructed to ensure the balance of task assignment.The multi-agent reinforcement learning algorithm is used to train the robotic arm to avoid collisions and find the optimal path when picking tasks.After the validation dataset test,the success rate of multi-manipulator autonomous path generation reaches 87%.Make a robot prototype for precise picking of high-quality tea sprouts,develop a remote monitoring and interactive system,and conduct functional tests.The speed control test and tracking accuracy test were carried out using the prototype,and the speed error was within 2.5%and the tracking accuracy error was within 1.5%,meeting the needs of precise tracking.In the picking test,the accuracy rate of bud picking can reach 90%,and the picking efficiency of double-arm is 85%higher than that of single-arm,picking a single sprout takes 0.86 s,which verifies the efficiency of multi-arm picking.Through the tracking and picking experiments,it is verified that the precision picking robot of the high-quality tea single bud developed in this paper meets the application requirements of the precise picking of the single bud of the robotic arm.