Fault Location Identification in Power Transmission Networks

This paper proposes a novel fault localization method that is based on the non-intrusive fault monitoring (NIFM) techniques in high-voltage/extra-high-voltage (HV/EHV) power transmission networks. In this work, the fault signals measured at the utilities can be extracted by the hyperbolic S-transform (HST). To carefully select coefficients of the HST representing fault transient signals and slash the size of inputs for recognition algorithms, power-spectrum-based HST is adopted in this paper to quantitatively transform the HST coefficients (HSTCs). After the processes of feature selection, the fault location indicator is recognized by the support vector machines (SVMs). To examine and validate the proposed methodology for constructing power transmission networks, various fault types are simulated by using electromagnetic transients program (EMTP). The simulation results are achieved to reveal that the proposed methods demonstrate a high success rate of fault location identification in power transmission networks for NIFM applications.