Research On Uncalibrated Robot Visual Servoing System

In order to overcome the drawbacks of the proposed method for calibration, uncalibrated method has been designed in a robotic hand-eye coordination process. Without calibrating the parameters of the camera and the robot in advance, the error signal is defined in terms of image feature parameters to design the controller, and driving robot. The system error converges to the allowed region of the robot hand-eye coordination.Based on the summary of research on the uncalibration robot of visual servoing, found the simulation model of fixed camera and eye in hand, and study the positioning of stated object and the tracking of moving. The establishment of a three-dimensional for positioning tasks of industrial robot visual servo system platforms-MOTOMAN-SV3XL industrial robot visual servoing control system. The main work in this paper is summarized as follows:1).Introduce the Support Vector Machine on statistical learning theory and the basic principles of Support Vector Machines, and make a brief introduction of the Least Squares Support Vector Machines (LS-SVM). A new approach to solve the inverse kinematics of robotic manipulators based on Least Squares Support Vector Machines (LS-SVM) is presented.2).Research the two link robot of uncalibration and complete 2D fixed target positioning and tracking moving targets. The proportion controller is used for visual positioning and visual tracking in visual feedback; The fixed camera and eye in hand structure of uncalibrated visual system is used in MOTOMAN-SV3XL.3).Construct robot visual servoing hardware system and program control software with VC++, carry out many experiments with the control strategy described above.4).With single eye in the hand of three-dimensional positioning experiment, a detailed description of the principle of image feature selection is given, the variables of image feature is confirmed, the corresponding visual mapping model also is proposed, visual positioning in three-dimensional space is successfully finished. Experimental results prove the validity of the algorithm.