Robust control for planar manipulators with image feature parameter potential

Abstract

In this paper, we deal with a planar manipulator control problem using camera information. We derive two types of the controller to compensate the full Lagrangian manipulator dynamics using the Lyapunov stability theory. One controller uses the image Jacobian, and the other controller requires the manipulator Jacobian and the direct kinematics. Hence these controllers are expected to have real-time computational advantages, and a robustness against the miscalibration of the camera intrinsic parameters, focal length, etc. In order to analyze the stability of the latter controller, we exploit the property of the rotational matrix and propose a new type of potential function of the image feature parameter space. The controller based on the image feature parameter potential has a nonlinear saturated potential term. Experimental results are presented to illustrate the controllers performance.