Faulty feeder detection of single-phase-to-ground network using improved morphological hybrid filter

Detection of single-phase-to-ground fault is a difficult task in power distribution network, especially in the Peterson-coil-grounded system, as the fault current is too small to be evaluated and the noise make it be difficult to extract the fault signal. In this paper, an improved hybrid filter (IHF) based on mathematical morphology (MM) is proposed to detect the faulty feeder in single-phase-to-ground network. Comparison studies with other MM-based filters show that the IHF has best performance of noise elimination. After the zero-sequence fault current is processed using IHF, the faulty feeder is detected by using the criterion of amplitude and phase. The proposed IHF can avoid the wave shift efficiently and is also good at removing the noise. The simulation results validate the merits and the effectiveness of the scheme.