Study On Kinematic Parameter Calibration For Industrial Robots

As a typical representative of intelligent manufacturing equipment,industrial robots cannot achieve high accuracy due to the errors of processing,assembly,and connecting rod deformation.This limits its transformation and upgrading to high precision and high intelligence.The measurement and identification of kinematic parameter errors are effective methods to improve robot accuracy,which have received widespread attention.Existing robot position measurement methods are difficult to meet the requirements of a large measurement range and high measurement accuracy,and the accuracy of the position and distance error identification model also needs to be improved.Thus,kinematic parameter calibration methods for industrial robots by measuring sphere center are proposed.The R-test prototype based on three orthogonal displacements and its structural parameter calibration method are designed.The method for identifying the robot kinematic error parameters that can be revised in the controller with the positioning accuracy and the distance accuracy is studied.The main contents are as follows:The forward and inverse kinematics of the vertical six-joint industrial robot are established.A kinematic parameter error model is established.The contribution of the error parameters that can be revised in the robot controller is analyzed by a simulation.Based on the performance evaluation indexes of the industrial robots' accuracy,the positioning errors and positioning distance errors are used as the dependent variables to identify the kinematic error parameters.According to the error range of the industrial robots,the target measurement accuracy and range of R-test are determined.Two measurement methods for industrial robots are proposed to meet different requirements.An R-test based on three orthogonal displacements is proposed.Orthogonal displacements make the measurement range distribution uniform.And a calibration method for non-redundant structural parameters is proposed,which further improves the measurement accuracy.The measurement method for obtaining three sets of displacements intermittently is proposed.A spherical contour error measurement method based on reference point calibration is designed,and the error of the sensor measuring spherical surface is accurately obtained.A kinematic parameter identification method based on the hand-eye pose parameter separation model is proposed.Based on the separation of hand-eye pose parameters,a kinematic parameter error identification model is constructed.And three independent parameters are used to represent the hand-eye rotation matrix.The least-squares and particle swarm optimization algorithm are used to realize the mixed step-by-step identification,and the hand-eye attitude parameters under the optimal kinematics parameter errors are obtained.To obtain the positioning errors of the identification,a position measurement method of the robot based on the same displacement is designed.A kinematic parameter identification method based on the accurate distance error model is proposed.The redundancy of the first joint angle error to the distance error identification model is deduced.An accurate identification model using distance errors with the global distance information is proposed.The first joint angle error is compensated to improve the relative position accuracy.To obtain the distance errors of the identification,two measurement systems for position based on the R-test are designed.The prototype of the R-test based on three orthogonal displacements,the position error measurement system based on codirectional displacement and R-test are built.The proposed identification method for the structural parameters of R-test,kinematic parameter identification method based on the hand-eye pose parameter separation model and the accurate distance error model are validated.Experimental results demonstrate the effectiveness and practicability of the proposed methods.