Transmission line model influence on fault diagnosis

Artificial neural networks have been used to develop software applied to fault identification and classification in transmission lines with satisfactory results. The input data to the neural network are the sampled values of voltage and current waveforms. The values proceed from the digital fault recorders, which monitor the transmission lines and make the data available in their analog channels. It is extremely important, for the learning process of the neural network, to build databases that represent the fault scenarios properly. The aim of this paper is to evaluate the influence of transmission line models on fault diagnosis, using constant and frequency-dependent parameters.