Enhanced segmentation of disturbance records by adaptive thresholding

dc.contributor.authorStachel, P.
dc.contributor.authorZivanovic, R.
dc.contributor.authorSchegner, P.
dc.contributor.conferencePower Systems Comutation Conference (16th : 2008 : Glasgow, Scotland)
dc.date.issued2008
dc.description.abstractTo analyse power system fault events captured by disturbance recorders the sampled analogue waveforms (voltages and currents) have to be segmented into system states (pre-fault, fault,...). In this paper a combination of wavelet transform (WT) and linear prediction coding (LPC) is applied to the symmetrical space phasor signals of the voltage and current waveforms. The signals obtained contain impulses when the system state changes abruptly. A threshold for segmentation is calculated using median filtering with adaptive window length. The segment borders are finally merged and validated.
dc.description.statementofresponsibilityStachel Philipp, Schegner Peter and Zivanovic, Rastko;
dc.description.urihttp://www.pscc2008.org/
dc.identifier.citationProceedings of the 16th Power Systems Comutation Conference, 14-18 July, 2008
dc.identifier.isbn9780947649289
dc.identifier.urihttp://hdl.handle.net/2440/54650
dc.language.isoen
dc.publisherUniversity of Strathclyde
dc.publisher.placeCD
dc.titleEnhanced segmentation of disturbance records by adaptive thresholding
dc.typeConference paper
pubs.publication-statusPublished

Files