Performance of an Elman recurrent network-based relay to determine the faul
t direction on a transmission line on fault data recorded in the field is d
escribed in this paper. The network uses samples of voltage and current sig
nals to learn the hidden relationship existing in the input patterns. Studi
es show that the proposed network is able to identify the fault direction c
orrectly and rapidly. Results obtained indicate that the proposed network i
s robust, fast and accurate.