Disturbance Extraction Methods for Applying Αpqm to Voltage Sag Localization in Distribution Network

This paper proposes three disturbance extraction methods (DEM) in order to apply αPQM, which is wildly deployed in distribution network, to voltage sag cause analysis such as localization and classification. The proposed DEMs overcome the drawbacks of the classical DEMs, which require a large PQM storage space and whose sampling frequency is preferably an integer multiple of the system frequency. For each DEM, a virtual steady-state waveform with sufficient length is constructed based on the pre-fault recordings whose length is no more than one cycle, which is subtracted from the during-fault recordings on the basis of aligning the phase angle to extract the disturbance. The SNR performances of the proposed DEMs outperform the classical ones in terms of fault duration, sampling frequency, residual voltage and antinoise. All of the DEMs are tested on the IEEE 123-bus distribution network and it has been demonstrated that the proposed approaches have better adaptability to voltage sag localization algorithm.