In this paper, an approach is proposed for accurate locating of single phase faults in transmission lines using voltage signals measured at one-end. In this method, harmonic components of the voltage signals are extracted through Discrete Fourier Transform (DFT) and are normalized by a transformation. The proposed fault locator, which is designed based on Random Forests (RF) algorithm, is trained based on these normalized harmonic components. RF algorithm has the capability of learning patterns with a large number of features. The proposed approach only requires voltage signals measured at one-end; hence, there are not problems of transmitting and synchronization of two-end data. In addition, current measurement is not required and the proposed approach is sheltered against current transformer errors and its saturation. No need for very high sampling frequency is another advantage of the proposed approach. Numerous tests carried out on a sample system indicate that accuracy of the proposed fault locator is secure against changing fault location, fault inception angle, fault resistance, and magnitude and direction of pre-fault load current. An average of 0.11% is obtained for the fault locating test errors.

Keywords