结合精度补偿的机器人优化手眼标定方法

A simultaneous calibration method based on Lie group theory and iterative optimization for robot orientation and hand-eye calibration is proposed to solve the problem of high precision vision measurement and position for industrial manufacturing field robots. The method uses optimization after enhancing the accuracy of robot. Firstly, the pose relationships between the vision system and the robot, that is, the hand-eye relationship and the spatial conversion relationship between calibration plate and robot coordinate system, are taken as the objects for optimization. The forward kinematics solution of the robot and the pose relationship between the vision system and the calibration plate are represented by a homogeneous coordinate matrix, and then a closed-loop optimization equation of the hand-eye relationship of the robot is constructed. Furthermore, a three-dimensional rotation group is used to represent a rotation matrix and to establish a calibration model equation. Then the nonlinear global optimization is used to simultaneously obtain initial solutions of rotation and translation of the matrix in the calibration equation. The calibration accuracy is improved by minimizing the reprojection error of the camera. Finally, the model accuracy of the robot is improved by using the robot kinematics calibration equipment, and more accurate calibration result is achieved by carrying out visual calibration. Experimental results show that the calibration method can be completed automatically by teaching a few points in advance, and the operation is simple and efficient. Precision analysis shows that the accuracy of the visual calibration algorithm is significantly improved from 0.15 mm to 0.1 mm by compensating the arm length and joint zero error of the robot body. Comparisons with the classical hand-eye calibration methods show that the proposed algorithm has the best calibration accuracy and stability in different test data sets.