Automatic Controller Tuning of Web Tension Control System via Online-type FRIT

Abstract

This paper proposes an online-type decentralized controller tuning method for the decentralized control system of a web transfer system, in which the mutual interactions are significant. The controllers are updated online by employing the fictitious reference iterative tuning (FRIT) and the online particle swarm optimization (OPSO) methods based on the structure of overlapping decomposition. It is demonstrated experimentally that the combination of FRIT and OPSO can improve the control performance with time using the test bench.