With the increasing application of industrial robots in production activities,the research and development of robot technology is becoming more and more important.Positioning accuracy is one of the most important performances of a industrial robot.The repetitive positioning accuracy of the robot has reached below 0.1 mm,but its absolute positioning accuracy is above 1mm or even 2-3mm,which cannot meet the needs of high precision off-line programming and intelligentization application.In order to improve the absolute positioning accuracy,many new calibration methods on kinematic parameters calibration are proposed.Based on the process of research parameter calibration,this article studies the key technologies in analyzing and improving the parameters.The main research and innovation ideas are as follows:1.Basic kinesiology theory of the robot end position and posture is researched.Kinematic algorithm is deduced and analyzed,then the simulation result indicates that the algorithm model is correct by the robotics toolbox.2.Aimed at the question that the laser tracker only measures the three-dimensional coordinates of the robot's terminal which may result in the attitude error of the calibration,a new six degree of freedom measurement method based on laser tracker is proposed and a new type of light pen is designed.Firstly,the coordinate system of the tracker is calibrated relative to the camera coordinate system.On the basis of coordinate measurement and attitude data acquisition,the measurement value of six degrees of freedom is calculated.The experimental results show that the position precision of this method is higher than 0.1mm in the range of 10 m,and the attitude precision is better than 0.01 degrees.Compared with the traditional monocular vision measurement and Leica T-probe,the position precision is better than the monocular vision method,and the position and posture precision is equal to the T-probe.It basically meets the needs of industrial precision measurement.3.Efficiency and precision of coordinate transformation between a robot and a laser tracker are improved.In order to unify the tracker's data and robot's data firstly,the method is proposed to calibrate the robot tool coordinate system,and on the basis of measuring the common points,we use the Rodrigo matrix to transform the coordinate system of the robot and the laser tracker.Finally,the experimental results are compared with the traditional methods.The result shows that the conversion error of the proposed method is limited to the positioning error of the robot itself;and it is more applicable to the special environment;and this method reduces the point position measurement,improving the conversion efficiency of the coordinate system.4.In order to improve the robustness and stability of parameter identification in the process of industrial robot calibration,a kinematic parameter calibration method based on robust ridge estimation is proposed.Firstly,the end pose error model is established based on the DH model,and the equivalent weight function is analyzed.Then the robot base coordinate system and the end position are measured by laser tracker.Finally,combining with the IGG3 weight factor function and the ridge parameter,the kinematic parameters are identified by robust ridge estimation.The experiment shows that the RMS error of the absolute positioning of the robot is reduced from 0.87 mm before compensation to the 0.21 mm after compensation.Compared with the traditional least square identification algorithm,this calibration method has higher fitting accuracy and stronger robustness and stability. |