Spreading the control complexity in decentralized control of interconnected systems (Q759691)

The problem is considered of stabilizing a large scale (dimension n) system using decentralized dynamic feedback laws. It is shown that the order of the i-th local controller can be taken to be at least the number of stable zeros of the i-th subsystem, while the total order of the decentralized controller does not exceed n-1. The decentralized controller introduces stable uncontrollable or unobservable modes into the overall system, thus lowering the effective order of that system. This enables the designer to spread the controller complexity between the subsystems. All the proofs are constructive so that conceptual algorithms for the design of the feedback laws may be derived.