Robust model predictive control of nonlinear processes represented by Wiener or Hammerstein models

Abstract

Representing nonlinear systems by linear models along with structured or unstructured uncertainties and applying robust control strategies could reduce the computational complexity in comparison with implementing the nonlinear model predictive controllers. In this paper design of robust model predictive controllers which are based on special classes of nonlinear systems representations called Wiener and Hammerstein are presented. The proposed algorithms approximate the nonlinear systems by uncertain linear models and reduce online the computational demands in the control implementation. The advantages of the proposed approaches are illustrated by two examples.