Research On Accuracy Analysis And Structural Parameters Calibration Of 6R Industrial Robot

In the processing of industrialization,the industrial robots as core equipment are required higher precision by more industrial fields.It is important to study the positioning accuracy of industrial robots.In this paper,6R industrial robot is studied on kinematics to research the effect of deterministic geometric parameter s and random joint gaps on the positioning error.According to its characteristic,the geometric error is chosen in the calibration.Through the closed-loop calibration method and the identification method based on the constraint model,the accurate geometric parameters are obtained to improve the robot accuracy.The workspace is got by geometric method and Monte-carlo method.Based on the positive and inverse kinematics,the influence of geometric parameters on the end accuracy is analyzed.Joint clearance is modeled using vector method.Based on theories of probability,the model between the second,third joints clearance and end positioning error is built.The resulting error distribution can effectively guide the robot's design,use and maintenance,ensuring its repeated positioning accuracy.Based on differential kinematics theory,the error transfer model of 6R series robot considering geometric parameters is established and the accuracy of the structural parameter error model is verified by Matlab simulation.The point,plane,sphere constraint methods in robot self-calibration are studied and a new sphere center distance calibration method is proposed.Newton method,least square method and Nelder-Mead algorithm are compared and analyzed to find the most suitable method in parameter identification.Considering the identifiability,17 identifiable parameters are determined,excluding unrecognized ones.Filtering the measuring poses,the optimal set is obtained by using up-down algorithm of DETMAX.The attitude information is involved to improve the accuracy of parameter estimation.The single sphere constraint calibration experiment is simulated.The parameters are estimated by Nelder-Mead method to tend the points to fit the sphere.It shows that the spherical model is feasible.Aimed at the robot self-calibrating without external measurement,two calibration schemes are designed based on plane and sphere center distance to test the validity.By collecting and processing the data points separately,the structural parameter deviation is obtained by Nelder-Mead method and compensated in controller.Experimental conclusions are obtained.