Zastosowanie sztucznych sieci neuronowych do lokalizacji zwarć łukowych w liniach elektroenergetycznych w warunkach nasycenia przekładników prądowych

The paper presents the method of arc fault location on power transmission lines using artificial neural network. The investigation was focused on the cases of CT saturation, resulting in the current signal transformation errors. Fault location is based on measurements of voltages and currents from two ends of the line, with knowledge of fault type. The accuracy of the location method was tested for different operating conditions of 400 kV power transmission line modeled in EMTP.