Faults recognition in power distribution systems using artificial intelligence
Abstract
In this paper, we will present a novel method to classify the short-circuit (SC) faults in power distribution systems using artificial intelligence. Four different types of SC fault including 3-phase, phase-ground (LG), phase-phase (LL) and phase-phase-ground (LLG) are randomly created in the IEEE 15 bus system using Matlab Simulink. The voltages and currents at the substation are captured to feed into the input layer, while the corresponding faults type are labeled in the output layer of the artificial intelligence models. 8000 data points are generated and randomly split into training and testing sets. The results show that deep learning algorithms can classify the short-circuit in the power distribution network with the accuracy of 95%.
Moraref