A robust velocity field control

Abstract

This paper is devoted to velocity field control (VFC) of uncertain robotic manipulators. We propose a proportional-integral (PI)-type controller derived from modeling error compensation ideas and singular perturbation theory, that requires a minimum knowledge of the plant (i.e., constant estimate of the inertia matrix). It is shown that semiglobal practical stabilization is achieved; that is, given any compact set of initial velocity field errors, there exist PI control gains which guarantee that the robot tracks a desired velocity field with arbitrary accuracy. The proposed controller was experimentally evaluated on a two degrees-of-freedom arm.