Identification and Compensation of Mechanical Errors for a SCARA Robot.

Abstract

This paper presents a new kinematic calibration method based on condition number and an error map method for SCARA robots. The geometric parameters can be identified more precisely by measuring the positions of a hand with minimum condition number configuration. Furthermore, the error map method improves the accuracy of geometric parameters. The experimental results demonstrate that the calibration method made the maximum absolute positioning error of a typical SCARA DD robot less than 0.25 mm all over the working area, while the maximum error was 2.20 mm in the case of the conventional method.