Finite Horizon Robust Model Predictive Control Using Linear Matrix Inequalities

Authors:	Chu Danlei, University of Alberta, Canada Chen Tongwen, University of Alberta, Canada Marquez Horacio J., University of Alberta, Canada
Topic:	2.1 Control Design
Session:	Advances in Predictive Control
Keywords:	Model predictive control, linear matrix inequalities, robust stability, terminal cost constraints

Abstract

In this paper, we develop a finite horizon model predictive controlalgorithm which is robust to model uncertainties. A moving average systemmatrix is constructed to capture model uncertainties and facilitate futureoutput predictions. The paper is focused on step tracking control. Usinglinear matrix inequality techniques, the design is converted into asemi-definite optimization problem. Closed-loop stability is treated byadding extra terminal cost constraints. The simulation results demonstratethat the approach can be useful for practical applications.