A novel approach for fault location on transmission lines

A new approach for power transmission line fault location is investigated. The fault locator estimation is based on the noise generated by the fault in the sending-end current signal. The proposed scheme consists of a wavelet-based filter module, a Prony-based signal processor and an ANN-based estimator. Input data has been generated using the Alternative Transient Program (ATP). A three-phase, frequency-dependent (FD) transmission line model was used. The scheme is tested using data employed in the ANN training as well as new data sets. The proposed locator has a good level of accuracy.