Constrained model predictive control: an experimental study

Constrained model predictive control (CMPC) has become increasingly popular in process industries in recent years owing to its optimization and constraint handling capabilities. This type of strategy features several aspects not commonly encountered in traditional advanced controls. Success with CMPC will be enhanced by an in-depth understanding of the underlying concepts. Training in CMPC therefore, has also become increasingly important in recent years. Such training is for the most part provided by off-line simulations. In this paper, the authors describe a real-time demonstration experiment to bring out the salient features of CMPC