Design of tension control systems to minimize interaction and disturbance propagation in web process lines

Abstract

In roll-to-roll systems (R2R), a local controller is designed for each section or tension zone based on measurement of web tension and speed from that section; this is typically referred to as decentralized control. This is particularly suitable for R2R systems because of ease of implementation and their structure. Since the material is transported from the unwind to the rewind through the process sections, the entire machine can be divided into several sections and decentralized controllers may be utilized for each section. It is important to understand the dynamics of web transport from one section to other downstream sections, that is, interaction between different sections, and how disturbances are propagated. Tension control system designs that minimize disturbance propagation will aid in improving process performance.

In this paper, we will first investigate minimization of interaction between subsystems of R2R systems when decentralized feedback and feedforward controllers are employed. In particular, we will consider a new interaction metric for dynamical systems which will quantify the amount of interaction; and show how model-based feedforward action can be gainfully employed to minimize disturbance propagation. We will also discuss control strategies that will minimize disturbance propagation. To evaluate the proposed designs and recommendations, we will show results from experiments conducted on a large experimental web platform.