Kohonen Neural Network Applied to Distance Protection of a Double-Circuit Line

This paper elaborates settings and modification of the operating characteristic of adaptive distance relay which protects double-circuit lines. The zero sequence mutual coupling of parallel circuits carry error into the relay impedance measuring. So there are two possible behaviors of distance relay: underreach and overreach depending on the operating mode. This paper proposes adaptive model which solves problems of single-line-to-ground fault on a single line. Input values are local measurements such as voltage and current plus the transmission line theory, and the output is signal for the distance relay whether it must trip or must not trip. As part of the adaptive model it is implemented self-organizing neural network - Kohonen network. The developed model offers solution for distance protection of a double-circuit transmission lines.