Research On Robot Hand-Eye Online Calibration Technology For Electronic Industry

With the development of science and technology,robot technology has made great achievement.The extensive application of industrial robots with integrated vision systems in various fields,especially,has become a hot issue for scientific research.Before obtaining target information through visual sensors,the robot requires to complete hand-eye calibrate.However,there are some problems with the current calibration methods,which cannot meet the requirements for accuracy and speed of electronic product production lines.Therefore,this thesis researches on the above-mentioned research on Hand-eye online calibration technology,and then realizes the rapid calibration of the hand-eye relationship of the robot.The research lays the research foundation for the robot to realize its tasks of movement and assembly in the actual environment.The main research contents are as follows:(1)In the study,the Eye-in-Hand hand-eye system with the camera mounted on the end effector of the robot is used to build the robot hand-eye system,the components and functions of the robot hand-eye system are analyzed and the camera's internal and external parameter matrix and distortion parameters are calculated based on the camera model.Providing evidence to robot hand eye calibration.Moreover,the kinematics model of the robot is established by the D-H method,and the kinematic relationship between the joint coordinate systems of the robot is given.(2)A black-and-white checkerboard calibration board is used as the target.For the Eyein-Hand robot system,an automatic calibration data acquisition algorithm based on block image histograms is proposed.The algorithm analyzes the orientation of the calibration board to provide the robot with a direction of movement,so that the robot can drive the camera to shoot the calibration board image according to the movement strategy,and extract the calibration data required for calibration,thereby,achieving the accurate shooting of the calibration board image and automatic collection of calibration data,and improving the automation degree of robot hand-eye calibration.The real-time online calibration process is of great significance to improve the calibration accuracy and calibration speed.(3)Based on the "two-step method" hand-eye calibration principle,an online feedback hand-eye calibration algorithm based on weighted least squares is proposed.According to the collected calibration data,the initial error is calculated as the weighted least squares weight.Then,solve the rotation matrix and translation vector of hand-eye relationship.On the one hand,reduce the influence of noise on the calibration accuracy,and on the other hand,reduce the transmission of error when solving the translation vector;It can judge whether the calibration result meets the calibration requirements in real time,so that the online calibration of hand-eye can adjust the calibration process according to the accuracy requirements.(4)In response to the production needs of the electronics industry,an industrial calibration experiment platform was set up using an industrial camera and a UR5 robot.The purpose of the experiment was to set a reasonable amount of calibration data within the tolerance range to achieve the optimal allocation of calibration accuracy and calibration time.The experimental results show that the hand-eye online calibration method proposed in this paper is feasible and effective.