Some Topics On Robot Kinematic Calibration

We focus on kinematic calibration based on computer vision in this thesis. Computervision based measuring system is set up. Camera is self-calibrated, an experiment isdesigned to evaluate the accuracy of the measuring system.We propose in this paper an improved geometric formulation of POE (Product OfExponential) based kinematic calibration of serial robots. We use both joint offset-free formulation and adjoint transformation errors of joint screws, and apply it to thecalibration of an elbow manipulator. Our formulation explains why the original POEcalibration always fails with the existence of joint offset errors; the adjoint formulationof joint screw errors eliminates joint screw constraints that was imposed in the originaliterated least square calibration algorithm.Another contribution of this paper is the proposal of a modified POE formulationwhich adopts point measurement data instead of frame measurement data of the end-effector, which can be more realistic and convenient for practical implementation.Simulation results show that the proposed method is plausible and effective.In the experiment part, the kinematic calibration of Orthopod manipulator based ongauge and computer vision is taken. To evaluate the simulation results, an experimentis builded up Adept Viper s850, the accuracy of the robot is improved from 0.635mmto 0.072mm.