Robust nonlinear tracking control of robotic manipulators

This thesis has discussed the development and implementation of robust nonlinear tracking control for a parallel and serial topology Tetrahedral robot (Tetrabot), although the theoretical control strategy presented is applicable to any manipulator tracking problem. The design of robust tracking controllers involves deriving a tracking law for uncertain dynamic systems, such that the actual positions closely track desired trajectories. Two new schemes, a robust sliding mode control and a Lyapunov-based robust tracking control, have been presented for uncertain dynamical systems in the presence of model uncertainty and disturbances. The foci of this study are the concepts and techniques of robust nonlinear tracking control with a bias toward industrial applications.;The Tetrabot system structure, hardware, software and the results of implementation on the three degree of freedom parallel geometry have been studied. In order to implement robust tracking control laws, the Tetrabot system software has been further developed.;Most importantly, the results of implementation of a nonlinear tracking controller on the Tetrabot rig facility are also studied. To demonstrate the performance attainable by this control strategy, the trajectory involved movement across the primary working volume to the end-effect point which is the largest distance possible and involved the continuous motion; such a motion will invoke a wide range of possible nonlinear dynamic representations. The proposed control strategy is robust to variations in robot loading. The experimental results obtained for the closed-loop response indicate that compensation, which employs explicit off-line parameter estimation, can improve tracking accuracy significantly. Using the robust tracking controllers, the position errors were smaller than those obtained using the original PID controllers. The robust tracking controller showed excellent results.