A study on the fault indentification of underground cable using neural networks

This paper presents a fault identification system based on neural networks for underground cable transmission systems (UCTS). EMTP was used for necessary transient data in training for fault type identification purposes. Data for various fault types in the underground cable system were generated and were used in training backpropagation neural networks. For the operation of the system a new data is tested for fair assessment of the designed system. Normalization of input data is adopted for reliable learning in neural networks. A proper size of the neural network was found via trial and error method, a brute-force method. This system was tested with various fault distances and fault incidence angles and proved its reliability.