Structural approach for the identification of power network dynamic equivalents

Abstract

A new method for deriving minimum-order dynamic equivalents of power networks is developed. The approach starts with a linearized model in polynomial matrix form of the external power network, seen from its boundary nodes with the study system. Then, the concept of structural dynamic equivalent is introduced as a generalized definition of equivalent related to the order reduction that may occur when deriving the transfer function matrix of the external system and due to pole-zero cancellations. In this case, necessary and sufficient conditions are given for determining systematically the existence of structural equivalents of external systems, their order and exact mathematical representation. Furthermore, the above-mentioned theoretical results are used to develop an algorithm for constructing approximate reduced models, called nearly structural dynamic equivalents, which takes into account the more realistic event of near pole-zero cancellation. Indications about the error bounds of the approximate model are also provided. The capability and usefulness of the proposed method are illustrated by carrying out simulation studies on a sample power network.