An adaptive precise one-end power transmission line fault locating algorithm based on multilayer complex adaptive artificial neural networks

This paper presents a new algorithm for precise fault locating in power transmission lines based on a novel two-layer complex adaptive linear neural network (CADALINE). The first layer is used for a direct estimation of the symmetrical components. The second layer is used to estimate the fault location directly from the time-domain positive sequence components of only one-end of the transmission line. The proposed algorithm is evaluated through extensive simulations in PSCAD/EMTDC and MATLAB software. The results presented in this paper clearly show that the CADALINE algorithm is highly robust and very accurate in performing the requested task of locating the fault.