Applications of Model Predictive Controllers in a Sugar Factory

Abstract

This paper presents two applications of Model Predictive Control in a sugar factory: temperature control in the diffusion process and density control in the wastewater treatment plant. The implementation is done using a Generalized Predictive Controller (GPC) designed for a wide class of industrial process, with the same computational requirements as a PID routine and embedded in the existing control system. The processes have in common the existence of long and uncertain dead times, therefore the original GPC algorithm is improved by the use of the T polynomial, which increases the stability robustness by filtering the predictions.