Design For The Motion Control Systems Of Manipulators Based On Iterative Learning

With the advancement and development of automated production process, as well as in-depth study of the robotic control technology, the application of industrial manipulators in the industrial automated production can reduce the costs of production and improve the efficiency of production. So the research and application of manipulators will become the future trend. And the research to design a motion control system of manipulators is bound to have a broad market prospect and research significance.In order to develop a motion control system of multi-joint manipulators with low-costing, low-power and high-performance, this dissertation mainly do the following work based on the analysis and research for the development of the industrial manipulator and its control technology.Firstly, based on the needs analysis for motion control system of multi-joint manipulators, and the prospect of market applications, a motion control solution based on PC and STM32was proposed. Then the STM32F407microcontroller as the core, and completed hardware design for the motion control module of multi-joint manipulator.Secondly, used D-H method to build the manipulator kinematic model and kinematic equations, analyzed and solved the forward and inverse kinematic problem of the manipulator. Then make the manipulator plane operating trajectory simulation on Matlab, which to verify the correctness of its kinematic solutions.Thirdly, we proposed an improved D-type iterative learning control law and proved its convergence after the theoretical analysis of the open and closed-loop D-type iterative learning control algorithm. Then as the manipulator as the research object, which is created the dynamics model for manipulator by Lagrange method, to simulate and analyze the iterative learning control algorithm.Fourthly, a motion control experiment of the manipulator was studied based on the PC plus STM32motion control module. This experiment used iterative learning control to achieve the trajectory tracking, and the results showed that it has a well actual tracking effect. Finally, the research work of this dissertation was briefly summarized, and the problems which need to be improved were proposed for the further research.