Propylene/Propane Distillation Colum Model Based Control - A New Approach

Abstract

Firsts model based control algorithms have used only linear process models, but as this
control methods were implemented and tested in some cases appeared robustness problems due to
process nonlinearities. Therefore, in order to eliminate this inconvenience, new control methods
based on the nonlinear model of the process were used but, the main disadvantage in this case is
the complexity of the obtained algorithm over the linear one. The main objective of this paper is to
bring some relevant contributions on model based control of a distillation process. Because the
process is a nonlinear one, with gains and dynamics that changes substantially with the operating
point, the main feature of the proposed model based control structures will be the ability to take
account of these nonlinearities. As a compromise between the linear and nonlinear model based
control approach, the model in which the proposed control algorithms rely will be a nonlinear
one, but represented as a reunion of linear models, one for each operating point.