Robust economic model predictive control: disturbance rejection, robustness and periodic operation in chemical reactors

Abstract

ABSTRACTThis study investigates the properties of a robust economic model predictive control (REMPC) algorithm with respect to rejection of disturbances in initial conditions and non-stationary disturbances, robust stabilization and closed-loop performance in the presence of model parameters’ errors, and the enforcement of optimal economic periodic operations. A key characteristic of the algorithm is that it enforces robustness to model errors without requiring terminal conditions (unless periodic operation is desired) and instead a set-point trajectory is calculated at each time interval. Robust stability and convergence to the calculated set-point trajectory are enforced online by a set of constraints. Three case studies are used to illustrate the closed-loop performance of the REMPC algorithm for reactor operation under different conditions. The algorithm is shown to preserve stability in the presence of model parameters’ mismatch. In the face of disturbances, the algorithm leads to higher profits when a variable set-point is used compared to a fixed one. Furthermore, the periodic operation obtained through the application of the robust algorithm confers an improved average cost compared to steady-state operation.