Estimation of Friction and States for Robust Control of Position Servo

Abstract

Friction is an inherent nonlinearity in electromechanical systems. It adversely affects the system performance. The key challenge is to get its accurate information as it is not measurable. This paper proposes a two-observer concept for friction and state estimation. Novelty of proposal includes finite time convergent estimations of the friction component and the system states. These estimations are used to design a robust control for accurate position tracking in servosystem. It is shown that accurate friction and state estimations are obtained, and control efforts are reduced while achieving precise performance. Tracking performance is improved using proposed control. The outcome of the proposed methodology is validated through computer simulations and experimentations.