Robust decentralized control of large-scale systems via H_∞ theory and using descriptor system representation

In this paper, a method for design of linear decentralized robust controllers for a class of uncertain large-scale systems in general form is presented. For a given large-scale system, an equivalent descriptor system in input–output decentralized form is defined. Using this representation, closed-loop diagonal dominance sufficient conditions are derived. It is shown that by appropriately minimizing the weighted sensitivity function of each isolated subsystem, these conditions are achieved. Solving the appropriately defined H_∞ local problem for each isolated uncertain subsystem, the interactions between the subsystems are reduced, and the overall stability and robust performance are achieved in spite of uncertainties. The designs are illustrated by a practical example.