A quadratic programming approach to constrained receding horizon H∞ control

This work is a modified version of an earlier work, which was based on ellipsoidal type feasible sets [Y.I. Lee and B. Kouvaritakis, 2000]. Unlike the earlier work, polyhedral types of feasible sets are adopted to deal with the input constraints. The use of polyhedral sets enables the formulation of an on-line algorithm in terms of QP (quadratic programming), which can be solved more efficiently than the semi-definite algorithm. A simple numerical example shows that the proposed method yields larger stabilizable sets with larger bounds on disturbances than the earlier approach [Y.I. Lee and B. Kouvaritakis, 2000].