A Rank Constrained LMI Algorithm for Decentralized State Feedback Guaranteed Cost Control of Uncertain Systems with Uncertainty Described by Integral Quadratic Constraints

The paper presents a numerical algorithm for constructing a decentralized state feedback guaranteed cost controller for an uncertain system. The uncertain systems under consideration contain structured uncertainty described by integral quadratic constraints. The decentralized controller is designed to achieve a closed loop system which is absolutely stable and such that a cost function satisfies a given bound for all admissible uncertainties. The main result gives an algorithm for constructing the desired controller in terms of LMIs subject to rank constraints.