Li, LiUgrinovskii, Valery AOrsi, Robert2015-12-100005-1098http://hdl.handle.net/1885/58229This paper addresses the problem of decentralized robust stabilization and control for a class of uncertain Markov jump parameter systems. Control is via output feedback and knowledge of the discrete Markov state. It is shown that the existence of a solution to a collection of mode-dependent coupled algebraic Riccati equations and inequalities, which depend on certain additional parameters, is both necessary and sufficient for the existence of a robust decentralized switching controller. A guaranteed upper bound on robust performance is also given. To obtain a controller which satisfies this bound, an optimization problem involving rank constrained linear matrix inequalities is introduced, and a numerical approach for solving this problem is presented. To demonstrate the efficacy of the proposed approach, an example stabilization problem for a power system comprising three generators and one on-load tap changing transformer is considered.Keywords: Constraint theory; Decentralized control; Feedback control; Linear matrix inequalities; Markov processes; Parameter estimation; Riccati equations; Robust control; Stabilization; Absolute stabilization; Discrete Markov state; Rank constrained linear matrix Absolute stabilization; Decentralized control; Markov jump parameter systems; Output feedback; Rank constrained linear matrix inequalities; Robust control; Uncertain systems200710.1016/j.automatica.2007.03.0162015-12-09