Fault Location Estimator Design for Power Distribution System Using Artificial Neural Network

Fault location in distribution system is critical issue to increase the availability of power supply by reducing time interruption for maintenance electric utility companies. estimator using artificial neural network developed line ground, line, ground and three phases faults system. To develop this one rural radial feeder Ethiopia, south west reign, Aba substation Tarcha used as a test feeder. This simulated ETAP software generate data different fault condition, with resistance loading conditions, which phase voltage current. The generated preprocessed put an input be trained. MATLAB R2016a toolbox train ANN programming graphic user interface estimator. feed forward multi-layer topologies improved back propagation, Liebenberg Marquardt learning algorithm network. After trained mean square error performance, regression plot histogram analysis was made found have excellent performance coefficient 0.99929 , validation 0.000102 range 0.015 0.019. In thesis practical implementation records at handled intelligent electronic device (IED) installed feeders. record IED can read PCM600 tool laptop or manually IEDs human machine interface, recorded estimate well type. Finally it that networks are alternate options design where sufficient available narrow distance from substation. has benefits assisting plan, saving efforts finding economic time. Keywords : Power distribution, intelligence, network, DOI: 10.7176/JETP/12-5-01 Publication date: November 30 th 2022