Self-calibration of a Stewart parallel robot with a laserranger

Abstract

In order to overcome influence of kinematic parameter errors on end-effector pose (position and orientation) accuracy of the Stewart parallel robot, a new self-calibration method with a laserranger fixed on the mobile platform is presented. Firstly, a new calibration scheme is proposed and the direct kinematic model of the parallel robot is built. Then the distance error model is derived, and the objective function of calibration is given. Subsequently, an improved differential evolution algorithm (DE) is adopted to identify the kinematic parameters. At last, simulation is made in a set of validation positions to verify the above calibration method. The results show that after calibration pose accuracy of the parallel robot has a significant improvement. At the same time, the strong stochastic search ability of DE is validated. © 2010 Springer-Verlag.