New Formulation of Dynamic Output Feedback Robust Model Predictive Control with Guaranteed Quadratic Boundedness

This paper extends a previous work for output feedback model predictive control (OFMPC) for nonlinear systems represented by a Hammerstein-Wiener model, to the case of dynamic output feedback. To assure asymptotic stability to a neighborhood of the origin, the present paper uses the notion of quadratic boundedness. The techniques for dynamic output feedback and quadratic boundedness allow a considerably different formulation of the controller, which is less conservative. Moreover, different from the previous work, the present paper assures recursive feasibility of the optimization problem. As a result, the main features for a synthesis approach of model predictive control (MPC) are restored for the problem studied in this paper. An example on the model of fuel cell system is given to illustrate the effectiveness of the controller.