An infinite horizon model predictive control for stable, integrating and unstable systems

Rafael Ribeiro Sencio¹; Darci Odloak¹

¹ Department of Chemical Engineering, Polytechnique School of the University of São Paulo

Baixar PDF doi:10.20906/CPS/CBA2018-0942

Resumo

Several works in the literature of model predictive control (MPC) have focused on the development of MPC formulations that are suitable for industrial applications. Usually, these controllers are based on state space models whose structures depend on the system type, which may be stable, integrating, or unstable. Thus, the usage of different internal models often yields distinct MPC formulations, which may be an issue for building a more general industrial package. Therefore, in order to consolidate these approaches, the present study addresses the development of a more general infinite horizon model predictive controller (IHMPC). This strategy is based on a novel formulation of state space model for stable, integrating, and unstable systems that is suitable for the IHMPC implementation. Simulation results demonstrated the successful application of the proposed controller to a deisobutanizer distillation column and to an unstable reactor system.

Palavras-chave: Model predictive control; MPC; Infinite horizon; Integrating systems; Unstable systems