ANN based fault location for medium voltage distribution lines with remote-end source

Abstract

This study presents a fault location algorithm for medium voltage (MV) overhead power distribution lines based on artificial neural network (ANN). In the study the possibility of connection of a small scale remote-end source connection to the end of a radial distribution feeder has been considered. In the study, feed forward ANN with back propagation algorithm with Levenberg-Marquardt training function is used. The ANN inputs were formed by using frequency information of fault data which were obtained with digital filtering. The algorithm is extensively tested for a various system conditions for the faults created on the overhead distribution system which has been modeled with Matlab/Simulink software. From the results attained it is seen that the proposed technique has not been significantly affected from the connection of a small scale hydroelectric generator to the existing distribution system.