Positional error similarity analysis for error compensation of industrial robots

• The positional errors of the industrial robot has similarity in the working space.
• The error similarity decreases as the distance between two positions increases.
• The proposed method doesn't need to identify or modify the kinematic parameters.
• The proposed method can effectively reduce the maximum positional errors of the robot.

The purpose of this paper is to propose an error compensation method with error similarity analysis to improve the absolute positional accuracy of industrial robots. The positional error similarity is proposed with the analysis of the error model established by robot kinematic parameters, and is quantified with semivariogram function. Then an error compensation method is proposed base-on positional error similarity. The ordinary Kriging method is used to calculate the positional errors of the robot TCP. The error compensation is performed by modifying the position coordinates in the robot controlling commands. Experimental verification showed that the maximum positional error of the robot TCP was reduced by 75.36% from 1.2912 mm to 0.3182 mm.