數(shù)學(xué)形態(tài)學(xué)在電力系統(tǒng)中的應(yīng)用研究

1、東北電力大學(xué)碩士學(xué)位論文數(shù)學(xué)形態(tài)學(xué)在電力系統(tǒng)中的應(yīng)用研究姓名：郭躍霞申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：電力系統(tǒng)及其自動(dòng)化指導(dǎo)教師：李天云20080301AbstractOnthebasisofanovelnonlinearsignalprocessingmethodcalledmathematicalmorphology(MM)，thispaperstudiedthepowersystempowerquality(PQ)disturbancesd

2、etectionandlocation，thepartialdischarge(PD)signalextracting，theleakagecurrentofmetaloxidesurgearrester(MOA)extractingandtheearlyfaultdiagnosisofMOAInsum，theworkinthispaperisasfollows：1BasedontheMMandthepoweralgorithmofsh

3、ortdatawindowanovelmethodofPQdisturbancesdetectionandlocationisproposedaccordingtothecharacteristicsofPQsignalsFirstlyinordertofilterrandomandimpulsenoises，asimpleandpracticalmorphologicalfilterbasedontheprincipleofMMisd

4、esignedItholdsthefundamentalcharacteristicsoforiginalsignalswhenfiltersnoisesandismoreefficientthanwaveletandtraditionalmorphologicalfiltersThen，tothefilteredresults，thedisturbancesalelocatedpreciselybypoweralgorithmofsh

5、ortdatawindowaccordingtothechangingenergyThevoltageswell，sag，interruptionandshortdurationfluctuationareusedtoverifyitsvalidity2AccordingtothecharacteristicsofPD，anewmethodusingempiricalmodedecomposition(EMD)andMMisapplie

6、dtoextractingthesignalsfromthePDsignalsburiedinexcessivenoisesItdosen’tneedtodefineanybasicfunction，SOithasstrongfilteringadaptabilityFortheintensiveperiodicnarrowbandwidthinterference，asimplemethodcombiningFFTandMMisput

7、forwardtosuppresstheinterferencebasedonthedifferencesbetweenthePDandinterference，nlesimulationandexperimentalresultsshowthattheEMDandMMmethodcallefficientlyremovethenoiseswithoutchangingthefundamentalshapeoftheoriginalsi

8、gnalsThistechniquesolvesthedifficult)rindenoisingandpreservingthefeaturesofsignalsAtthesametime，themethodcombiningFFTandMMCaneliminatethenarrowbandwidthinterference，andthedistortionofPDsignalsiSverylOW3Inordertoextractth