Chaotic Characteristic of Time Series of Partial Discharge in Oil-Paper Insulation

在线阅读 下载PDF 导出详情
摘要 Thechaoticcharacteristicsoftimeseriesoffivepartialdischarge(PD)patternsinoil-paperinsulationarestudied.TheresultsverifyobviouschaoticcharacteristicofthetimeseriesofdischargesignalsandthefactthatPDisachaoticprocess.Thesetimeserieshavedistinctivefeatures,andthechaoticattractorsobtainedfromtimeseriesdifferedgreatlyfromeachotherbyshapesinthephasespace,sotheycouldbeusedtoqualitativelyidentifythePDpatterns.Thephasespaceparametersareselected,thenthechaoticcharacteristicquantitiescanbeextracted.ThesequantitiescouldquantificationallycharacterizethePDpatterns.TheeffectsonpatternrecognitionofPRPDandCAPDarecomparedbyusingtheneuralnetworkofradialbasisfunction.Theresultsshowthatbothofthetworecognitionmethodsworkwellandhavetheirrespectiveadvantages.Then,boththestatisticaloperatorsunderPRPDmodeandthechaoticcharacteristicquantitiesunderCAPDmodeareselectedcomprehensivelyastheinputvectorsofneuralnetwork,andthePDpatternrecognitionaccuracyistherebygreatlyimproved.
机构地区 不详
出版日期 2011年06月16日(中国期刊网平台首次上网日期,不代表论文的发表时间)
  • 相关文献