Control of linear systems subject to input constraints: A polynomial approach. (Q5932276)

The paper deals with the local stabilization of a linear multi-input multi-output discrete-time system originally described in the state space subject to hard input constraints by a dynamic controller. The initial state is assumed to belong to a given convex subset of the state space. Polyhedra and ellipsoids are considered. Two different problems are studied. First, find a stabilizing dynamic controller for the system subject to input constraints when the initial state belongs to a given polyhedron or ellipsoid. Second, find a dynamic controller and a stability region that is as large as possible such that the system subject to input constraints is stabilized when initiated within this region. Easily implementable algorithms for computing a stabilizing control law and enlarging the stability region are presented. They are based on the Youla-Kučera parametrization and geometric properties of polyhedra and ellipsoids casting the above problems into standard convex programming problems solved by linear second-order cone and semidefinite programming.