Research On Control System Of 5-DOF Motion Platform Based On Vision

Robot visual servoing is a hot subject in the field of machine vision. It involves many element areas including robotics, control theory, image processing and so on. Vision robot has the ability to analyze environmental information and to perform tasks independently, so it is widely used in the application of aviation, national defense and industry automation. In the paper, visual servoing control system based on the image is studied with the five degree of freedom(5-DOF) motion platform, and each part of the system is detailedly analyzed and designed. After that the feasibility of the system is proved by MATLAB simulation and experiments. The main contents of the research are as follows.(1) Design visual servoing control system of the five degree of freedom motion platforms. The system includes upper computer software, motion control card, servo driver, servo motor, camera and proximity switches, etc.(2) Build the kinematics model of the five degree of freedom motion platform, and obtain the solution of its positive and inverse kinematics. Image Jacobian matrix representing the relationship between the velocity of robot in the world plane and the velocity of the target in the image plane is derived. Then the motion control law is designed and its stability is verified by Lyapunov method. In the end, two simulation experiments including fixed target location and moving target tracking based on MATLAB are carried out to verify the validity of the system model and the motion control law.(3) Using Kalman filter instead of the traditional analytical method to identify image Jacobian matrix can avoid the huge amount of calculation and the calculation error of the Jacobian matrix inversion, which is verified by the simulation experiment.(4) Vision module is the key of visual servoing system. Based on the visual pattern of eye-in-hand, establish the pinhole model of the camera and calibrate the internal and external parameters of the camera by OpenCV. Then collect image information of the target by the camera and analyze it through the technology of digital image processing. Finally, image coordinate of the target is obtained by target recognition technology.(5) Control software of the system includes upper computer software, Pewin32PRO2 real-time control software and communication between the two and PMAC. The upper computer software including control module and visual module is programmed with VS2010 and OpenCV and parameters is set on the real-time control software. Communication between the two is the method of memory file mapping.(6) Two experiments including planar graph contour engraving and moving target tracking are carried out with the 5-DOF motion platform, which are respectively corresponding to the two simulation experiments in the third chapter. From the process of experiments, we can see that the visual servoing control system of 5-DOF motion platform can effectively complete the target localization and tracking with good tracking accuracy and real-time requirement, although the movement of the motion platform is relatively slow.