Feature-based High Impedance Fault Detection in the Distribution System Using Hilbert and Fourier Transforms

This paper presents a new technique based on the combination of Hilbert and Fourier transforms for addressing the problem of high impedance faults (HIFs) detection in electrical distribution systems. HIFs are not efficiently detected by conventional protection systems because of their low current magnitudes. The proposed HIF detection method is as follows: First, perform the Hilbert transform on each phase current to obtain the instantaneous phase signal. Second, apply the Fourier transform to obtain the spectrum of these instantaneous phase signals and the instantaneous phase of synthetic pure sinusoidal signal to calculate the error spectrum. HIF index is calculated as sum of the error spectrum. Finally, HIF is detected using change point detection (CPD) technique on the HIF index. A large number of simulations in the PSCAD/EMTDC software on the IEEE 123-node test feeder and field data tests on different HIFs and non-HIFs cases show that the method has high detection and low false positive rates.