POSITIONING CONTROL OF A HUMAN-MACHINE COOPERATIVE GRAFTING MANIPULATOR FOR UNSTRUCTURED ENVIRONMENTS

摘要

Crown grafting of fruit trees has the disadvantages of high labor intensity and reduced graft survival. Therefore, a human-machine cooperative manipulator that relies on passive joint braking was designed to realize position control. The manipulator can replace manual operations to solve the problem of different positions in the grafting process and provide positioning and force support for canopy grafting. This study determined that the working space of the manipulator can cover the canopy area of fruit trees. Dynamic equations were established for motion simulation and feedforward compensation control of the manipulator. According to the dynamic model, the joint braking process was simulated. The simulation results showed that the joint braking torque needs to be dynamically controlled to ensure positioning accuracy of the manipulator. A process of passive joint braking was designed based on the proposed ideal braking curve. By comparing the position control accuracy of independent proportional integral derivative (PID) control, dynamic model feedforward compensation control, and PID control based on feedforward compensation of the dynamic model in simulations, it was determined that PID control based on feedforward compensation of the dynamic model was suitable for application in the braking torque control system. Finally, prototype tests showed that PID control based on feedforward compensation of the dynamic model can realize high-precision joint braking and position control of the manipulator. The positioning error was less than 5%, and the maximum vibration acceleration amplitude was reduced by 26.7% to 68.5%. The control system of the manipulator, using PID control based on feedforward compensation of the dynamic model, can provide adaptability for unstructured environments and reduce power consumption for application in field operations.