Finite-time velocity sensorless integral sliding mode control for roll-to-roll systems under matched disturbances.

Abstract

Roll-to-Roll (R2R) systems operating in complex manufacturing industries encounter challenges from inherent system factors like model uncertainty or external interference. These factors directly lead to a significant reduction in productivity and output quality of the system, thereby causing great economic losses. In response to the growing demand for consistently high performance in industrial R2R processing, we propose a finite-time integral terminal sliding mode control integrated with a finite-time extended state observer. The proposed controller achieves consistently high tracking performance of transport speed and material surface tension of the system owing to the integral terminal sliding mode control method. Meanwhile, the extended state observer is integrated with three component observers to address the dual estimation task for velocity-based dynamic models of unwinder roll, active guide roll, and rewinder roll, encompassing both velocity states and disturbances/uncertainties. In this manner, observational errors under the estimation error dynamic models are proved to be ultimately uniformly bounded by the Lyapunov theory. Furthermore, the control framework, including an integral sliding controller and an observer, is also demonstrated through the Lyapunov stability theory. Comparisons with other nonlinear controllers are conducted and discussed to validate the efficiency of our proposal.