New Reconfiguration Scheme for Distribution Networks Using MLP Neural Networks

Abstract

Recently, the advantages of using distributed generation have been analyzed in most of scientific discussions. Distributed generation rapidly changes
the configuration of the electric industry from monopoly to competitive and will totally change the configuration of power systems in the 21st
century. One of the main changes, that can be a disadvantage of distributed generation presence, is changing the radial configuration of distribution
networks. Injection of DGs currents to a distribution network results in losing radial configuration and consequently losing the existing coordination
among protection devices along with decreasing reliability of the system. In this paper, a new method for distribution system reconfiguration has
been presented which by implementing on distribution networks including DG, in addition to maintaining the efficiency of their protection system;
optimizes DGs operation through islanding operation and causes the system reliability and total ENS of the system to be decrease. The suggested
method has been implemented on a sample distribution network using DIgSILENT Power Factory 13.2 and its results have been presented.