Measurement of and Reverse Compensation for Spatial Position Error of Industrial Robots

Abstract

Abstract A measurement system was designed to meet high positioning accuracy requirements for trajectory-determined industrial robots in its partial workspace. Positioning error was calculated and its varying laws were analyzed by tracking the spatial position of the tool center point (TCP) of the industrial robot using a laser tracker and then transforming the relevant coordinate system. A neural network model of the difference between the commanded TCP position and the positioning error of the industrial robot was built. A reverse compensation model of the positioning error of the industrial robot was constructed based on the reverse compensation principle of positioning error, after which the commanded TCP position coordinates were revised accordingly. The experimental results show that the proposed method is feasible and effective, the average positioning error of the robot decreased to 0.028 from 3.232 mm, the standard deviation decreased to 0.100 from 2.987, and the positioning accuracy increased by more than two orders of magnitude. The proposed method is suitable for situations that involve determined trajectory, high positioning accuracy, and sealed robot control systems, and is easily incorporated into engineering applications.