Comparison of artificial neural networks and conventional algorithms in ground fault distance computation

This paper gives a comparison between an artificial neural network method and a differential equation algorithm and wavelet algorithm in transient based earth fault location in the 20 kV radial power distribution networks. The items discussed are earth fault transients. Signal pre-processing and the performance of the proposed distance estimation methods. The networks considered are either unearthed or resonant earthed. The comparison showed that the neural network algorithm was better than the conventional algorithms in the case of very low fault resistance. The mean error in fault location was about 1 km in the field tests using staged faults, which were recorded in real power systems. With higher fault resistances, the conventional algorithms worked better.