Method For Improving Positioning Accuracy Of Robot Based On PACMM

Due to the effect of error sources such as machining,The industrial robot has poor position accuracy,and its application scope is limited.However,with the concept of "made in China 2025" put forward,aerospace manufacturing and other fields put forward higher requirements for the accuracy of the industrial robot.In view of the low efficiency of the existing measuring instruments,this paper develops a Portable Articulated Coordinate Measuring Machine(PACMM)with variable arm length for robot calibration to realize positioning error compensation.The main research contents are as follows:(1)The kinematics model of industrial robot is established.The structural characteristics and working mode of UR5 industrial robot is analyzed,and the kinematics model of UR5 industrial robot based on MDH model is established.The influence of structural parameter error and non-structural parameter error on the accuracy is further analyzed,and the kinematics error model is established.(2)The Support Vector Regression Machine(SVR)method for robot error prediction is proposed.The derivation process of SVR method is introduced,and the error model of industrial robot based on SVR is established.By establishing the model scheme between the robot joint angle and the robot positioning error,the positioning error of industrial robot can be predicted.(3)A PACMM with variable arm length is developed for the industrial robot calibration.The kinematics model of the PACMM with variable arm length is established,and the error correction of each joint is realized.The whole calibration of the PACMM with variable arm length is realized by using LM algorithm.Finally,a high precision instrument with a positioning error less than 0.03 mm is obtained as a robot calibration equipment.(4)Experimental verification of robot positioning error compensation based on PACMM is carried out.The robot calibration experiment platform is built,and the prediction and compensation experiment of robot positioning error is realized by SVR method,and the experimental results are analyzed.The correctness and effectiveness of the SVR for the absolute positioning error compensation of the robot are also verified by the experimental results.