Double-circuit line adaptive protection based on Kohonen neural network considering different operation and switching modes

This paper deals with problems connected to the distance protection of double-circuit transmission lines. Because of the zero sequence mutual coupling of parallel circuits, the relay "sees" incorrect impedance. In this paper, the authors develop an adaptive model which solves problems of single-line-to-ground faults on a single line. Different operation modes change short circuit currents (I/sub k1/ - phase-to-earth short circuit current and I/sub k3/ - three-phase short circuit current) on both ends. In fact, the parameters of the active network will be changed. The idea was to develop an adaptive model with different independent Kohonen networks. Each network is trained with short-circuit examples according to the switching mode and operation mode. Input parameters are voltage, current, switching mode and operation mode. The developed adaptive model was tested on a double-circuit 400 kV transmission line.