Application of simple cascade GPC with robust behaviour to a sugar refinery

Abstract

This paper presents the application of a Generalized Pre dictive Controller (GPC) to sludge density control in a sugar factory. The loop is controlled by a cascade strategy where both the master and the slave controllers are predic tive ones. The control law is extremely simple to compute and the tuning is straightforward since a method to im plement GPC previously developed by the authors which is very simple to implement and tune has been used. The controllers are embedded in the existing control system needing the same computational requirements as pid rou tines. The original GPC algorithm is improved by the use of the socalled T polynomial which increases the stabil ity robustness by ltering the predictions in order to cope with model uncertainties and di erent process dynamics caused by changes in the process operating points