Robust decentralized control using an alternating projection approach

Abstract
Provides a geometric formulation of the robust decentralized control problem for both continuous-time and discrete-time systems. The problem is formulated as a feasibility problem of finding a set of matrix parameters in the intersection of a set of linear matrix inequalities (LMIs) and a non-convex rank constraint. Alternating projection methods are proposed for solution. Fixed-order, interval and structural constraints can be imposed on the controller in the same framework.