Robust Adaptive Bilateral Control of Teleoperation Systems with Uncertain Parameters and Unmodeled Dynamics

Kinematic and dynamic parameters of robot manipulators are difficult to measure exactly. Also, time varying unknown dynamical parameters of human arm, during interaction with the master robot and unknown parameters of environment during interaction with the slave robot, in teleoperation systems, insert further uncertainties. Furthermore, unknown parameters, unmodeled dynamics of master/slave robots, human arm model and environment introduce more uncertainties. In this paper, a robust adaptive master-slave teleoperation control strategy is introduced which require neither the exact knowledge about the parameters of the master/slave robots, human arm and environment, because of the adaptability with the unknown parameters, nor the exact dynamical equation of master/slave Falcon robot, because of the robustness against the unmodeled dynamics. Two Novint Falcon robots are used as master/slave robots and due to having the highly nonlinear complexity of these robots, they are considered as a single translational equivalent mechanism with known inertia, damping and stiffness coefficient and an unmodeled dynamic term because of this approximation. The human arm and environment are modeled as a 1-DoF mass, spring and damper model (MSD) with unknown coefficients. Moreover, an additional nonlinear spring and nonlinear damper has been used for better approximation of nonlinear property of the human arm and the environment. A Lyapunov function is introduced for stability and the position tracking convergence proof of the entire teleoperation system. The validity of the theory is confirmed by simulations.