Economic performance tracking for nonsquare MPCs based on a two‐layer approach

Abstract

Range tracking for nonsquare systems is frequently adopted in process industries and its implementation is developed by model predictive control technologies. However, these approaches differ from those most considered in academia, with set-point tracking and the same number of controlled and manipulated variables. In this scope, real-time optimization (RTO) emerges as a diffused technology to improve the economic performance considering process, safety, and environmental constraints. In this work, a way to integrate economic aspects in industrial model predictive controllers (MPCs) by treating the nonlinear economic function as an output of the process model is proposed. The tracking error of the cost function is monitored, and its real value is estimated by a state estimator. The approach was applied to a nonsquare range system, exemplifying a continuous stirred-tank reactor (CSTR) with Van de Vusse kinetics, and showed that it is capable of tracking the minimum cost operation robustly. The paper also compares the proposed strategy to the traditional RTO implementation, which provides optimized targets, and presents slight improvement in the steady-state operation regarding the optimal cost seeking.