MPC-based controller for augmented boiler-NOx model

This paper addresses the problem of controlling a drum type boiler operation as well as its NO_x emission level. For that purpose an augmented model integrating both boiler and NO_x models has been developed. The system has multiple roots at the origin causing poor phase and gain stability margins. In addition, one of its inputs has been considered as an undesirable disturbance. Different control strategies have been studied and benchmarked. A two-stage Augmented Model Predictive control (MPC) controller with constraints demonstrated best performance. Computer simulations show the performance of such approach.