| With the rapid development of society,the scale of industrial production continues to expand,so that industrial robots are widely used and developed.It is an important direction of robot development in the future that visual feedback information is added to robot motion control,which can increase the robot's perception of the environment and improve the robot's application range.However,the accuracy of the robotic visual servo system with visual information as feedback is restricted by many aspects,such as the positioning accuracy of the robot,the extraction accuracy of the object by the visual system,the calibration accuracy of the visual system,and the tracking accuracy of the robot visual servo system.Therefore,this paper studies the factors that affect the accuracy of robotic visual servo system,and verifies the accuracy of visual servo system of SCARA robot and the correctness of the design theory and method through the tracking experiment of moving objects on the conveyor belt.The main contents are as follows(1)The control scheme of visual servo system is designed,and the experimental platform of robot,camera and conveyor is built.Then,the main factors that affect the accuracy of visual servo system are analyzed,including robot positioning error analysis,visual system calibration error analysis,as well as the servo control accuracy analysis of moving target(2)Robot calibration and error compensation mechanism based on vision.In order to improve the absolute positioning accuracy of SCARA robot,an automatic calibration method of SCARA robot based on vision is proposed.Firstly,according to the D-H model of SCARA robot,the kinematics model of SCARA robot is established.Secondly,the collected image is preprocessed and the center pixel coordinates of the robot's end mark are extracted by the shape-based template matching method,then the position error of the robot's end is measured and corrected to get the corrected robot joint coordinates.Then,the parameters of robot error are identified based on the least square method.Finally,the arm length and zero point which affect the positioning accuracy of SCARA robot are compensated,and the compensation results are analyzed.(3)Automatic hand-eye calibration method of robot.Firstly,the camera model is analyzed,and the transformation relationship among coordinate systems is solved.Then the camera calibration is carried out to obtain the camera internal and external parameters,so as to reduce the system distortion.Then aiming at the problem of large calibration error caused by manual teaching robot coordinates in hand-eye calibration,A hand-eye calibration method based on mark point recognition is designed.Any mark point adsorbed on the robot end is obtained by vision and the transformation matrix between the mark point and the robot end coordinate is calculated,so as to establish the mapping relationship between camera and robot and complete automatic hand-eye calibration.The results of hand-eye calibration are analyzed.Finally,the coordinate relationship between conveyor belt and robot is established by three-point method,and the coordinate relationship between camera,robot and conveyor belt is established.(4)The realization of precise control of visual servo system.Aiming at the problem of low time delay and control frequency of visual servo system,a visual servo control method based on interpolation and Kalman filter is proposed.Firstly,Kalman filter is used to filter and predict the position,velocity and acceleration of the observed target position,and then the target position is interpolated in the observation period to improve the output of frequency of visual servo system and visual servo control effect of the robot.Then,the correctness of the design theory and method is verified by the experiments of tracking and absorbing the objects moving at constant speed and variable speed on the conveyor belt. |
PDF Full Text Request