LAUSR

We present a systematic framework to derive model-based simultaneous strategies for the integration of scheduling and control via multiparametric programming. We develop offline maps of optimal scheduling actions accounting for the closed-loop dynamics of the process through a surrogate model formulation that incorporates the inherent behavior of the control scheme. The surrogate model is designed to translate the long-term scheduling decisions to time varying set points and operating modes in the time scale of the controller. The continuous and binary scheduling decisions are explicitly taken into account in the multiparametric model predictive controllers. We showcase the framework on a stand-alone three-product continuous stirred tank reactor, and two reactors operating in parallel.