A Model Based Observer-Controller Formulation for Parametrically Uncertain Robotic Systems

This work proposes a novel robust observer-controller formulation for a class of second order nonlinear systems where only the output measurements and partial knowledge of the dynamical system parameters are available. Specifically, an observer formulation that makes use of nominal values of the dynamical system parameters in conjuction with a robust output feedback controller is designed to ensure semi-global exponential convergence of the position tracking error to a practical bound in the neighborhood of the origin. The stability of the closed-loop system and convergence of the tracking error terms are guaranteed via Lyapunov based arguments. Extensive numerical studies are presented to illustrate the performance and feasibility of the observer-controller couple.