A noise equivalent bandwidth approach to obtain reduced order models for power system excitation control

In this paper the concept of noise equivalent bandwidth is used to obtain reduced order models of power system excitation control over a specified frequency range. The criteria for measuring the efficacy of the reduced model, consists of maintaining the fidelity of the dynamic behavior of the system variables in time domain. For this reason the performance of the reduced model is compared with that of the exact model via simulation on a 3-machine power system. The reduced order models are used when excitation control is incorporated in the transient energy function method of direct transient stability assessment of power systems. The reduced order models provide a fairly accurate representation of the detailed models and reduce computational burden. This feature of the reduced order models is of significance in on-line security analysis of power systems.