Research On Robot Grasping Moving Object Based On Dynamic Visual Servo

Robot grasping moving object is to complete tracking and capturing moving object base on visual servo control. It is a cutting-edge application subject for the intelligent robot, and it has a good application foreground in the region of industry, spaceflight and entertainment. The complication of robot grasping moving object provides a good foundation for achieving more complicated tasks.Compared with the conventional sensor robot control, there are many distinct advantages of the visual servo control, more flexible, more precision. The visual servo has the strong robust for the robot calibration error. In dynamic visual servo, the robot acceleration and joint's torque are calculated based on vision detection result to realize the force servo of machine vision. The paper studies the robotic dynamic visual servo control. The realizing of grasping moving object task is faster and more flexible by dynamic visual servo. The dynamic visual servo is also optimized in the end.The essential works of the thesis can be summarized as follows:1. The 3D position of moving object is located based on dual stereo vision technology. After the relation of camera coordinate system and world system is calibrated, the visual servo information of moving object is filtered and pre-estimated by Kalman filter.2. The composed cell module of robot is analysised. The dynamic and static precisions of the cell module are also discussed, and some common methods to promote precision are also presented. The geometry structure relationship and force structure relationship are analysised, and the robot control model is visually simulated by OpenGL based on robot control system model of kinematics and dynamics.3. The method to optimize robot target trajectory is promoted, and the trajectory in joints coordinate system is gotten after decoupling. The robot arm can reach the moving object by time-optimal fuzzy Bang-Bang control. When the joint position close enough to the expectation value, a fuzzy tracking control with an integral item is adopted.4. The robot dynamic visual servo control is researched, and the position visual impedance control method is introduced to grasping moving object. With CMAC network compensator, the output of vision impedance control is optimized. Finally, the simulation of control experiment of grasping moving object is carried out.