Stability Comparison of Fault Location Techniques Against Faulty GPS Time Tagging for Three-Terminal Lines

Abstract

Fault location techniques based on the arrival times of Traveling Wave (TW) need precise time tagging and synchronization for multi-terminal transmission lines. Although this need can be satisfied by using Global Positioning System, extracting the correct arrival time of the main TW is a troubling task, especially for multi-terminal lines with large number of wave reflections in different directions after fault incident. So, stability of the proposed techniques should be studied when the time tagging has also faced faults. In this paper, three fault location techniques based on TW, Genetic Algorithm (GA) and Artificial Neural Network (ANN) are discussed and then their stabilities against the faulty time tagging are compared. Simulation results show that three discussed methods have relatively close error averages. However, the proposed ANN-based method is twice much more stable than TW- and GA-based methods in about 22.95% of samples.