Research On Kinematic Parameter Identification Of One Key Self Calibration Of Manipulator

The six degree of freedom manipulator is widely used in industrial robots.With the development of intelligent manufacturing technology,the application scenarios of industrial robots have been expanding to the fields of automobile assembly,electronic product assembly,medical auxiliary equipment,etc.,which require higher positioning accuracy,and also put forward higher requirements for the positioning accuracy of mechanical arm.At present,the calibration method is often used to improve the positioning accuracy of the manipulator.In the calibration process,the measurement link is the most important,usually relying on highprecision measurement equipment,requiring professional and technical personnel to operate.The calibration process is complex,requiring a large site and space,which is not easy to carry out in the working environment of the manipulator.The error model establishment and identification method of the calibration manipulator is also suitable for The parameter error of the solution has great influence.In this paper,a one key self calibration method based on plane constraint is proposed,which uses high-precision plane constraint to ensure the measurement accuracy,uses monocular vision instead of manual operation to simplify the measurement steps of calibration.At the same time,the kinematic modeling method of the manipulator,the establishment of the error model,the removal of redundant parameters in the error model,the selection of the identification position,etc.are studied in-depth in the research group On the basis of MATLAB simulation program,the robot developed by ourselves is tested.This paper studies and establishes the kinematic model of the manipulator.In view of the confusion of two commonly used D-H parameter modeling methods in the modeling process,it combs the process of the two modeling methods,summarizes the differences between the two D-H parameter models,and gives the skills of fast resolution of the two models.The differential transformation of D-H parameter model is carried out,the relationship model between the end pose error and D-H parameter error of the manipulator is established,and the reason of redundant parameters in the error model in the previous research results of the research group is analyzed.Then the reason of redundant parameters may appear in other conditions is analyzed in depth,and the method of removing redundant parameters in various cases is put forward,and the method of removing redundant parameters is deduced The error model after the additional parameters is modified according to the plane constraints.Finally,the modified error model is verified by simulation in MATLAB.In the calibration process,monocular vision is used to locate the calibration block.In view of the situation that the traditional gradient operator cannot extract the dark edge,the K-means clustering method in HSI image space is used to extract the dark edge,and the edge extracted by the traditional operator is supplemented to obtain the complete image edge.The corresponding relationship between the calibration block and the image is used to solve the camera and image Through programming,the robot can lock the position of calibration block by itself with the help of monocular vision,thus controlling the robot to measure the calibration block in contact.In the aspect of the selection of the identification configuration,five indexes are summarized to evaluate the advantages and disadvantages of the identification configuration,and three ideas are put forward to select the optimal configuration,which are single parameter analysis,single configuration analysis and combined configuration analysis.Experiments are designed to select the optimal configuration by using three analysis ideas,and finally the combined configuration analysis method is selected after experimental verification and comparison In the experimental verification process of the combined configuration analysis method,the artificial bee colony algorithm of the nonlinear optimization algorithm is selected to verify the correctness of the combined configuration analysis method in MATLAB,and the optimal identification configuration combination is obtained.On the basis of the previous research of the research group,this paper studies the one key self calibration method of the manipulator based on the plane constraint,provides a method to calibrate the manipulator automatically,and makes contributions to the use of the manipulator in higher accuracy,more efficient calibration,etc.