Design of a robust power system controller with constraints

Abstract

Power system stabilizers (PSSs) are generally used to damp low frequency oscillations. Usually, a classical PSS uses a fixed, second-order model, for which a mere tuning of its parameters is required. These controllers have been implemented successfully for many years. However, conventional PSSs may become inadequate when the operating conditions of the system are changing widely, because of the lack of robustness. For this reason, robust control method, specifically H∞ optimal control, has been proposed and applied to the design of PSSs. A characteristic of H∞ controller is that the order of the synthesized controller is high. This makes the controller complex and its implementation difficult. For easy implementation, it is desirable to obtain a low-order H∞ controller. One way to cope with higher order controller is to use a model reduction technique to reduce the controller's order. However, this may lead to some performance degradation. In this paper, the desired order of the controller is used as a constraint at the outset of the design process. The obtained low order controller is compared with a standard H∼ optimal controller which order has been reduced using conventional reduction techniques. It is shown that the controller obtained by imposing constraints at the design stage performs better than the one obtained using conventional reduction techniques.