Trajectory Tracking of Manipulator Based on Iterative Learning Control

Abstract

With the development of technology of automated production and robot, robots, which are at high levels of automation, are widely used in all aspects of production. The application of manipulator in production enhances the production efficiency and product quality, and slash the production cost. The research and design of manipulator motional control system, with low cost and high performance, has very important research value and significance. The iterative learning control is particularly effective for the repetitive round-trip motion control system, which is looking forward to meet the requirements of rapid response and the repeatability of the manipulator's work and then the manipulator reaches the predetermined space position and completes the given work content. This paper takes the double-joint manipulator as the research object, analyzes its kinematics by Lagrange dynamics equation, and establishes the kinematics model of the manipulator. Besides, the iterative learning control was applied to the tracking of the manipulator, and its convergence was tested. Three learning methods of d-type iterative learning control, improved d-type iterative learning control and pd-type iterative learning control are compared. And during this experiment the Simulink model was established. The final results of simulation manifest that the improved pd-type iterative learning control has better convergence and control accuracy than the previous two methods.