Adaptive Control Of Uncertain Nonlinear Teleoperation Systems

Kinematic parameters of a robotic manipulator are hard to measure precisely and the varying size and shape of tools held by the robot end-effector introduce further kinematic uncertainties. Moreover, the exact knowledge of the robot nonlinear dynamics may be unavailable due to model uncertainties. While adaptive master-slave teleoperation control strategies in the literature consider the dynamic uncertainties in the master and the slave robots, they stop short of accounting for the robots’ kinematic uncertainties, which can undermine the transparency of the teleoperation system. In this paper, for a teleoperation system that is both dynamically and kinematically uncertain, we propose novel nonlinear adaptive controllers that require neither the exact knowledge of the kinematics of the master and the slave nor the dynamics of the master, the slave, the human operator, and the environment. Therefore, the proposed controllers can provide the master and slave robots with a high degree of flexibility in dealing with unforeseen changes and uncertainties in their kinematics and dynamics. A Lyapunov function analysis is conducted to mathematically prove the stability and master-slave asymptotic position tracking. The validity of the theoretical results is verified through simulations as well as experiments on a bilateral teleoperation test-bed of rehabilitation robots.