Image-based adaptive visual servoing of 2D robots under Jacobian and dynamic friction uncertainties: theory and experiments

A globally convergent visual feedback control scheme is proposed for dynamical planar robot arms subject to uncertain camera, robot, analytical Jacobian and dynamic friction parameters. When complex friction arises, visual servoing suffers to drive the robot to the desired trajectories, in particular in slow motion and velocity reversals, which are typical motion regime in visual servoing due to the vision system properties. Moreover, dynamic friction is usually neglected in motion control and it is not the exception in visual servoing literature. In order to prove the theory described in this paper, the real-time OS, Linux-RTAI, is used to obtain experimental results of this controller on a direct-drive robot manipulator. Results suggest its excellent performance.