Detection of Fault in a 400 kV Transmission Line Using EMTP and ANN Method

Abstract

Power transmission and distribution networks are important for the supply of quality power to the consumers. The rating of the protective device and the co-ordination play a vital role in improving the power quality in Electrical power systems. Since, generating stations exists far away from the load centers and the transmission lines have to run over long distances, the probability of occurrence of faults in transmission lines are very high. Due to such line faults, the system gets destabilized and requires immediate isolation. Therefore, fault analysis is very critical and essential issue in electrical power system with regard to fault clearance and power supply restoration with minimum interruption. In order to detect fault in transmission line, simulation of fault conditions has been carried out by dividing the total length of transmission line into large number of π sections. The π section comprises the series inductance and resistance of line and capacitance to ground. Single phase, double phase and three phase to ground faults are simulated at various distance from generator transformer and fault currents are calculated using electromagnetic transient programme (EMTP). Fault currents are also calculated by simulating faults at different points on wave to account for randomness of fault occurrence. Various time ~ current characteristics are used to train neural network and tested for detecting a specific fault using the untrained data of several faults at other distances. A comparison was made for efficiency of fault detection between time domain and frequency domain data. The neural network calculation was extended to currents determined from point- on wave switching for fault simulation.
Copyright © 2014 Praise Worthy Prize - All rights reserved.