One layer gradient-based MPC+RTO of a propylene/propane splitter

ODLOAK, DARCI; HINOJOSA, ALDO IGNACIO; FERRAMOSCA, ANTONIO; GONZÁLEZ, ALEJANDRO HERNÁN

COMPUTERS AND CHEMICAL ENGINEERING

PERGAMON-ELSEVIER SCIENCE LTD

Lugar: Amsterdam; Año: 2017 vol. 106 p. 160 - 170

0098-1354

Here,the implementation of the gradient-based Economic MPC (Model Predictive Control)in an industrial distillation system is studied. The approach is an alternativeto overcome the conflict between the MPC and RTO (Real Time Optimization)layers in the conventional control structure. The study is based on the therigorous dynamic simulation software (SimSci Dynsim) thatreproduces the real system very closely and is able to communicate with Matlabwhere the control/optimization algorithm is implemented. The gradient of theeconomic function, which is required to the on-line execution of the extendedcontrol strategy, is obtained through the sensitivity tool of the real-timeoptimization package (SimSci ROMeo®). It is shown that the proposed integration approach leads toconvergence and stability to the closed-loop system. In order to study the prosand cons of new strategy, a propylene distillation system is simulated withboth, the proposed approach (one layer MPC+RTO) and the conventional two-layershierarchical structure of control and optimization. The results show that, fromthe performance, stability and disturbance rejection viewpoint, the proposedgradient-based extended control method for this particular system issubstantially better than the conventional approach.