Decentralized optimal control of a 2n-area power system

This paper considers the design of a fixed structure decentralized controller for a 2n - area power system. The controller structure allows that each local controller has the local state variables and the interaction variables available to it. The controllers consist of constant feedback gain matrices. The criterion for the specification of these gain matrices is the minimization of a centralized quadratic cost functional. This optimization is achieved via the solution of a set of necessary conditions which take the form of a set of simultaneous nonlinear algebraic equations. A computational algorithm is presented for the solution of the necessary conditions and simulation of the closed loop system (with 2n = 4) is used to illustrate the effectiveness of this controller under various types of operating conditions. This controller is compared to other controllers of the same structure as well as to the centralized optimal controller. Some of the distinctive features of the decentralized controller are delineated, and system trade offs are discussed.