Design Of Visual Servo Control System For Flexible Assembly Robot

With the integration and development of information technology and control theory,industrial robots have become indispensable automation equipment in contemporary industrial production,which improves product quality and work efficiency.However,there is still a need to improve the intelligence and flexibility of robot technology.The application of robot visual servo control technology to image sensors can improve their environmental perception and response ability to cope with different production work.In this paper,the KUKA KR5 arc industrial robot is used as the research object to construct the visual servo control system,and the robot is guided to complete the grasping and combination of the work piece objects.Firstly,this paper studies image processing algorithms in robot visual servo system,including image gray and contrast changes,spatial filtering and edge detection.For the speed and accuracy requirements of image processing,the image acquisition and processing flow of visual servo control is designed.Through the comparative analysis of edge extraction algorithms,the edge of target image is completed.The defect of large computation and long running time in the normalized cross correlation matching algorithm is improved,and the image pyramid layer by layer search concept is proposed to speed up the matching speed,and the calculation equation of work piece posture is deduced.Secondly,the visual servo control mode of the robot is analyzed,and the kinematics model of the robot is established by Denavit-Hartenberg method.According to the requirements of the system,the kinematics analysis of the manipulator is carried out.The forward kinematics analysis determines the functional equation of the joint rotation angle and the position and posture of the end of the robot.The inverse kinematics analysis solves the rotation angle of the current position of the manipulator corresponding to each joint,and analyses the trajectory planning problem of the robot.The methods of camera calibration and hand-eye calibration are studied.A visual system model is established.The transformation relationship between space coordinates and image coordinates is determined.The calibration error is improved.The effect of tangential component of the distortion model on the overall distortion model is taken into account to make the calibration more accurate.Finally,an experimental platform of visual servo control system for articulated robots is built.Aiming at the motion control part of industrial robots,the double closed-loop control system is designed.The simulation experiment is completed in Simulink using PID control algorithm.The experimental design and analysis results show that the visual servo control system improves the real-time performance of image processing and reduces the calibration error.The expected goal has been achieved.