Design For Robot Visual Servoing Controller

Robot visual servo control is a method to use the camera to extract image features as a control signal, drive the robot to complete the task of servo control. In the process of realization of visual servo control system, robot has to overcome the model of uncertainty factors, improve the control precision of the system, so as to improve the intelligent level of industrial robots, widening the application range of industrial robots.Due to the problems of robot controller in the visual servo control system. In this paper, the author designs a kind of switching controller based on cartesian coordinate and polar coordinate, and design the visual controller of using predictive control theory. Using the RBI-S6-10L robot as a platform, design of the industrial robot visual servo control system and the related experiment. Main works and achievements are as follows:1. Analyses the problems existing in the current robot visual servo system, based on the robot visual servo control system structure, the relevant mathematical model was established, building visual servo simulation model in the Matlab/Smulink platform.2. Due to the visual servo system using2D visual information, then the image information is lack of the depth. So the image target rotate180degrees around the camera optical axis, can produce mechanical arm robot back. According to mechanical arm robot problem, we design a controller to solve rotation problems, the controller based on polar coordinate3. The controller based on polar coordinate is less effective in dealing with a target translational motion and able to deal effectively with the target rotation problem. Using a traditional visual servoing structure and controller based on the polar coordinate, propose the switch controller based on the theory of the switch control system. Using switch controller can handle the translation and rotation of the target at the same time.4. The jitter and delay problems in the switch controller based on visual servo system, designed a controller based on model predictive control of vision, through the online rolling optimization dynamic visual servo control system of optimal image control signals. According to the RBI-S6-10L six degrees of freedom robot, the robot visual servo control system were designed in physical experiment platform By the RBI-S6-10L robot operation scenarios static target for the real-time positioning and fetching the experiment.