Abstract
Hexapods are increasingly being used for high-precision 6-DOF positioning applications such as for positioning mirrors in telescopes and for positioning samples in synchrotrons. These robots are designed and controlled to be very repeatable and accurate. However, structural compliance of these positioning systems limits their positioning accuracy. As accuracy requirements become more stringent in emerging applications, compensating for inaccuracy due to structural compliance becomes necessary.In this regard, firstly, a method for elastostatic calibration of hexapods is presented. This method uses a lumped stiffness parameter model to parametrize the relationship between the platform deflections and the force/moment applied on it. These parameters can be estimated using deflection measurements performed using known forces/moments applied on the platform. The estimated parameters can then be used to predict and correct hexapod’s positioning errors due to compliance.Secondly, a new approach is presented to optimize stiffness identification for robot elastostatic calibration. In this, a framework is proposed to formulate criteria to choose best set of poses and forces for stiffness identification experiment. The parameters identified under experimental conditions (poses and forces) suggested by these criteria ensure minimum impact of errors influencing stiffness identification (uncertainty of deflection measurements and errors in forces applied) on compensation quality. Additionally, it also maximizes accuracy after compensation at desired pose(s), along desired axe(s) of the platform and with desired forces/moments on the platform. This stiffness identification optimization framework ensures best compensation for positioning errors due to compliance as per the positioning requirements of the application at hand.Lastly, a method is presented to eliminate the influence of thermal deflection of a hexapod on the measured 6-DOF pose of its platform. This method is necessary when thermal deflections of the hexapod are large enough to impact results of a study, which was the case with some tests performed to validate methods developed in this thesis.The efficacy of presented methods have been validated by means of simulation studies on a bipod and experimental studies on a high-precision hexapod positioning system.