Robust predictive control for uncertain nonlinear MIMO systems based on MISO Volterra expansion on generalized orthonormal bases

Abstract

This paper proposes a new robust predictive control for uncertain nonlinear MIMO systems based on a set of MISO submodels where each is modeled by the 2nd order MISO-GOB-Volterra model, the coefficients of which belong to an orthotopic parameter uncertainty set (PUS). We present the general form of a new predictor and so, we propose an optimization problem formulated as a quadratic programming (QP) under linear and nonlinear constraints with respect to parameter uncertainties. The efficiency of the proposed multivariable robust predictive control approach is validated on an experimental Communicating Two Tank system (CTTS) considered as Two-Input and Two- Output (TITO) system.