Controle preditivo robusto baseado em modelo aplicado a sistemas não-lineares incertos linearizados por realimentação de estados

Abstract

In this work the problem of controlling a class of nonlinear uncertain dynamical systems is revisited based on a Model Predictive Control (MPC) approach. The fundamental ideal in the proposed methodology is to consider an uncertain linear model associated with the attempt of feedback linearizing the nonlinear uncertain system, while taking into account the original nonlinear systems input constraints. Polytopic uncertainties are employed to encompass both parametric uncretainties in the nonlinear system model and remaining mismatches related to the non-exact feedback linearization. Finally, a Robust MPC (RMPC) controller synthesis procedure is outlined relying on usual relaxed Linear Matrix Inequalities (LMI) conditions that shows superior performance when compared to standard LMIs.