Power system equivalent based on an artificial neural network

Abstract

Very often, insufficient data is exchanged between neighboring power systems for quality load flow and contingency analysis. The external systems, therefore, have to be substituted with the power system equivalents. In this paper the possibilities of using an artificial neural network as the external power system equivalent is explored, to be used for load flow and contingency analysis within the internal power system. The experiment is performed on a standard IEEE 24-node network which is, for the purposes of testing, divided into two systems (the internal and the external) and the external system is modeled by a neural network. The results are presented and discussed.