Multivariable Adaptive Control of the Rewinding Process of a Roll-to-roll System Governed by Hyperbolic Partial Differential Equations

In this paper, an active control scheme for the rewinding process of a roll-to-roll (R2R) system is investigated. The control objectives are to suppress the transverse vibration of the moving web, to track the desired velocity profile, and to keep the desired radius value of a rewind roller. The bearing coefficient in the rewind shaft is unknown and the rotating elements in the drive motor are various. The moving web is modeled as an axially moving beam system governed by hyperbolic partial differential equations (PDEs). The control scheme utilizes two control inputs: a control force exerted from a hydraulic actuator equipped with a damper, and a control torque applied to the rewind roller. Two adaptation laws are derived to estimate the unknown bearing coefficient and the bound of variations of the rotating elements. The Lyapunov method is employed to prove the robust stability of the rewind section, specifically the uniform and ultimate boundedness of all of the signals. The effectiveness of the proposed control schemes was verified by numerical simulations.