Research On Trajectory Tracking Of Apple Picking Manipulator

Apple picking is a labor-intensive operation,and as China’s urbanization continues to grow,along with an aging population,it has resulted in a huge labor shortage for apple picking.To address the existing labor shortage and decrease associated manufacturing costs,it can be solved by vigorously developing apple picking machinery.Industrial robots have been widely used and are becoming more and more mature.Due to the complexity of the agricultural operating environment and the higher requirements for robots,the development of robot applications in the agricultural field has been relatively slow.Apple picking manipulator is a typical intelligent agricultural equipment,integrating manipulator,sensor,controller and other advanced technologies.The picking manipulator must have precise joint and end position positioning,and accurate trajectory tracking to greatly increase position precision.Therefore,this paper investigates the trajectory planning and trajectory tracking of the picking manipulator using apple picking as the aim.The main work of the paper is as follows:(1)Kinematics and dynamics analysis modeling of apple picking manipulator.Taking the AR3 manipulator as the object,the forward kinematics model of the manipulator is established,and the inverse kinematics solution is solved by analytical method.The correctness of the model is verified by MATLAB simulation.The Lagrange method is used to model the dynamics of the manipulator.In order to simplify the model,the dynamic characteristics of the last three joints are ignored,and the feasibility of the simplified dynamic model is verified in the SIMULINK environment.(2)Research on trajectory planning of apple picking manipulator.Aiming at the trajectory planning problem of manipulator under the constraint of joint activity,the trajectory planning of polynomial interpolation algorithm is studied.Considering velocity smoothing and acceleration change rate jump,the trajectory planning of manipulator joint space and Cartesian space is realized by quintic polynomial interpolation algorithm.Based on the AR3 manipulator model,the comparative experiment in MATLAB environment shows that the quintic polynomial interpolation algorithm can effectively ensure the smoothness of joint angular velocity and acceleration.(3)Research on trajectory tracking of apple picking manipulator.Aiming at the slow response speed of sliding mode control,this paper improves the exponential reaching law.A fuzzy adaptive sliding mode controller based on disturbance observer is designed for external disturbance.In the SIMULINK environment,the controller of this paper is compared with the exponential reaching law controller.According to the simulation results,for the position and speed tracking of the manipulator joint,compared with the exponential reaching law control method,the proposed method can track the predetermined expected trajectory within 0.8 s,and the end trajectory tracking response is faster and more accurate.When the joint moves to the connection inflection point,the control performance of the controller is more stable and the chattering phenomenon of the system is reduced.(4)Experiment and result analysis.Aiming at the uncertain factors existing in the manipulator system in the actual working environment,based on the simulation experiment,the manipulator experimental platform system is built,and the structure of ’ host computer +industrial computer + AR3 manipulator ’ is adopted.The end trajectory tracking comparison experiment is carried out in the experimental platform system.The experimental results show that the maximum end trajectory tracking error of the controller in this paper is 2.62 mm,which is much smaller than the maximum error of the controller based on fuzzy adaptive,and meets the accuracy requirements of actual apple picking.