During the process of manufacture and measuring to industrial robots,the connection parts will deform because of long time working and loading,it will lead to the change of kinematic parameters of industrial robots,which further on leading to the decrease of positioning accuracy of industrial robots' extremity.The essay focuses on the optimization to positioning accuracy of industrial robots' extremity through two methods.The first method is to process off-line calibration to industrial robots.The second method is to estimate industrial robots' extremity motion parameters in order to build the base to online compensation to industrial robots.Regarding the first method,the essay applies laser ranging three dimensional network built by laser tracker to process the measurement instead of adopting laser tracker directly.Only adopting the length information from laser tracker makes up the lack of angle accuracy to laser tracker.This essay compares the measurement and compensation results of two methods,with the compare of positioning accuracy of industrial robots' extremity by application of laser ranging three dimensional network and laser tracker,the average error decreases from 0.309 mm to 0.246 mm,down by 20%.Regarding the second method,the essay applies the estimation of industrial robots' extremity motion parameters by using Untracked Kalman Filtering.Untracked Kalman Filtering can process nonlinear motion,meanwhile it will not eliminate higher order term error,and it has strong adapting ability to environment.Therefore,this method is used for general robot and measuring instrument.When industrial robots' extremity processing uniform linear motion,the location average error decreases from 0.159 mm to 0.088 mm because of the application of Untracked Kalman Filtering,down by 44.65%. |