Adaptive force control of robots in presence of uncertainty in environment

Environmental uncertainty is an important factor in performance degradation in most force control schemes. This paper deals with force and motion control of robotic manipulators interacting with a general class of environments whose both geometry and stiffness are not exactly known. Using only measurements of robot joint motion variables and the contact force, the proposed adaptive controller generates a torque control signal ensuring asymptotic tracking of desired force and motion trajectories. Identification of environment characteristics is carried out by three dynamic and one static estimator. Structure of the proposed controller permits its straightforward integration into most existing force control schemes dealing with robot dynamics uncertainty. Simulation examples are presented to illustrate the results.