文档名：基于反射系数与核函数构建的新型电缆缺陷诊断方法
摘要：为解决现有的电缆局部缺陷检测方法中缺陷定位效果差、缺陷类型判断困难等问题,提出一种基于反射系数谱(RCS)的新型电缆缺陷诊断方法.首先,利用电缆线路的分布参数模型,建立了含局部缺陷电缆线路的RCS数学模型,根据RCS数学模型,确定合理的核函数,设计出基于RCS的新型缺陷诊断函数,该缺陷诊断函数的极值峰可用于定位电缆线路中的缺陷,并确定缺陷的类型、严重程度等特征;然后,对不同类型与严重程度的电缆线路缺陷进行建模仿真诊断,测试该方法的可行性与准确性;最后,在同轴电缆与10kV电力电缆上设计了局部缺陷,并用该方法对真实电缆缺陷进行诊断分析.仿真与实测结果表明,该方法能精准地定位电缆线路中的缺陷,并且可以准确地判断出缺陷的类型与严重程度,说明了该方法的实用性,且具有较高的实际工程价值.
Abstract：Inrecentyears,asitsexcellentmechanicalandelectricalproperties,cross-linkedpolyethylene(XLPE)powercableiswidelyusedintheconstructionoftheurbanpowersystem.Withtheincreaseinservicetime,variouscablelocaldefectsoftenoccurduetothehostileoperatingenvironmentandpoorproductquality.Thelong-termdevelopmentofthedefectmaycauseacableinsulationfailure,whichseriouslythreatensthestabilityoftheurbanpowertransmissionanddistributionsystem.Therefore,itisofgreatsignificancetoaccuratelydiagnosecablelocaldefectsasearlyaspossible.Recently,byusingthebroadbandimpedancespectrum(BIS)andreflectioncoefficientspectrum(RCS),aseriesofcabledefectlocationmethodsbasedonfrequencydomainreflectometryhasbeenproposed.However,theexistingmethodshavesomeproblemsinengineeringapplications,suchaslowreliabilityofthepositioningresultandtheinabilitytodeterminedefecttype.Tosolvetheaboveproblems,thispaperpresentsanewcabledefectdiagnosismethodbasedonRCS.ByestablishingthedefectdiagnosisfunctionbasedonRCSandkernelfunctionconstruction,itcanaccuratelyobtainthecharacteristicsofthedefect,suchaslocation,typeandseverity.Firstly,basedondistributedparametermodelofthecableline,theRCSmathematicalmodelofthecablelinewiththelocaldefectisestablished.TheRCSiscloselyrelatedtothepositionandcharacteristicimpedanceofthelocaldefect,soitcanbeusedtodeterminethecharacteristicsofthelocaldefect.Then,accordingtothemathematicalmodelofRCS,areasonablekernelfunctionisdesignedinthispaper,andanewdefectdiagnosisfunctionbasedonRCSisobtained.Bytheperiodicityofthetrigonometricfunction,theproposeddefectdiagnosisfunctioncanformapeakatthepositionofthedefect.Finally,thepeakcanbeusedtodeterminethecharacteristicsofthedefect,suchaslocation,typeandseverity.Whenthecharacteristicimpedanceofthedefectisgreaterthanthatofnormalcable,thecorrespondingpeakispositivepolarity.Onthecontrary,whenthecharacteristicimpedanceofthedefectissmallerthanthatofnormalcable,thecorrespondingpeakisnegativepolarity.Inthispaper,twoclassiccabledefectmodelsareestablished.InthedefectmodelⅠ,thereisseriesdefect-resistanceinthecablecoreorshieldinglayertosimulatethedamagedefectofthecablecoreorshieldinglayer,leadingtotheincreaseofcharacteristicimpedance.Inaddition,thehigherthedefect-resistancevalue,themoreseriousthedefect.InthedefectmodelⅡ,thereisparalleldefect-resistancebetweenthecablecoreandshieldinglayertosimulatethedamagedefectofcableinsulation,leadingtothedecreaseofcharacteristicimpedance.Atthesametime,thelowerthedefect-resistancevalue,themoreseriousthedefect.Inthesimulationprocess,foursimulated250m10kVXLPEpowercablesamplesarebuilt,andallthedefectsarelocatedat200m.ThetypesofNo.1～4simulatedcablesamplesarethedefectmodelⅠ,Ⅰ,ⅡandⅡ,respectively.Whilethedefect-resistancevaluesare60?,120?,60?and120?,respectively.Simulationresultsshow:(1)Thedefectpositioningerrorsinallsimulatedcablesamplesare0.2m(0.1%),whichprovesthattheproposedmethodcanaccuratelylocatethedefect.(2)InthedefectdiagnosisfunctionsofNo.1,No.2simulatedcablesamples,thepeaksatthedefectarepositivepolarity.However,forNo.3,No.4simulatedcablesamples,thepeaksatthedefectarenegativepolarity.Asaresult,thepolarityofthepeakcanbeusedtodeterminethetypeofdefect.(3)AsthepeakabsolutevaluesofNo.1～No.4simulatedcablesamplesare799.1,1112.8,446.7and254.7,respectively.ThepeakabsolutevalueoftheNo.2simulatedcablesampleisgreaterthanthatofthe1#simulatedcablesample.Ontheotherhand,thepeakabsolutevalueoftheNo.4simulatedcablesampleissmallerthanthatoftheNo.3simulatedcablesample.Sothepeakabsolutevaluecanbeemployedtoassesstheseverityofthedefect.Inthemeasurementprocess,theexperimentalsubjectsare40mSYV50-5-1communicationcoaxialcableand105m10kVXLPEpowercable.TheNo.1～No.3measuredcablesamplesarethecommunicationcoaxialcable,whiletheNo.4measuredcablesampleistheXLPEpowercable.Assameasthesimulationresults,theproposeddefectdiagnosisfunctioncanaccuratelylocatethedefect,inwhichthemaximumpositioningerrorisonly0.7%.Atthesametime,thetypeandseverityofthedefectcanbedeterminedfromthedefectdiagnosisfunction.Theconclusionsofthisarticlearesummarizedasfollows:(1)Thepeakofthedefectdiagnosisfunctioncanbeusedtoobtainthelocationandtypeofthedefect.Forthepeakofthedefectdiagnosisfunction,theabscissavalueisthepositionofthedefect,andthesign(positiveornegative)ofordinatevalueisusedtodeterminethetypeofdefect,inwhichthepositivesignandnegativesigncorrespondtodefectswithincreaseanddecreaseofcharacteristicimpedance,respectively.(2)Thedamagetothecablecoreorshieldinglayercausestheincreaseofthecharacteristicimpedanceatthedefect,andtheproposeddefectdiagnosisfunctionformsapositivepeak.Thedamagetothecableinsulationcausesthedecreaseofthecharacteristicimpedanceatthedefect,andtheproposeddefectdiagnosisfunctionformsanegativepeak.(3)Intheproposeddefectdiagnosisfunction,thepeakabsolutevalueisrelatedtotheseverityofthedefect.Thehighertheabsolutevalue,themoreseriousthedefect.(4)Thesimulationandmeasurementresultsshowthattheproposedmethodcannotonlyaccuratelylocatethedefectinthecable,butalsodeterminesthetypeandseverityofthedefect.
作者：饶显杰 徐忠林 刘翔宇 关惠方 周凯 Author：RaoXianjie XuZhonglin LiuXiangyu GuanHuifang Zhoukai
作者单位：国网四川省电力公司成都供电公司成都610041四川大学电气工程学院成都610065
刊名：电工技术学报
Journal：TransactionsofChinaElectrotechnicalSociety
年，卷(期)：2024, 39(7)
分类号：TM41
关键词：电缆线路 局部缺陷 反射系数谱 缺陷定位 缺陷分析
Keywords：Cableline localdefect reflectioncoefficientspectrum defectlocation defectanalysis
机标分类号：TM726.4F205F127
在线出版日期：2024年4月12日
基金项目：国家自然科学基金基于反射系数与核函数构建的新型电缆缺陷诊断方法[
期刊论文] 电工技术学报--2024, 39(7)饶显杰 徐忠林 刘翔宇 关惠方 周凯为解决现有的电缆局部缺陷检测方法中缺陷定位效果差、缺陷类型判断困难等问题,提出一种基于反射系数谱(RCS)的新型电缆缺陷诊断方法.首先,利用电缆线路的分布参数模型,建立了含局部缺陷电缆线路的RCS数学模型,根据RC...参考文献和引证文献
参考文献
引证文献
本文读者也读过
相似文献
相关博文
基于反射系数与核函数构建的新型电缆缺陷诊断方法 A New Cable Defect Diagnosis Method Based on Reflection Coefficient and Kernel Function Construction
基于反射系数与核函数构建的新型电缆缺陷诊断方法.pdf
- 文件大小:
- 1.16 MB
- 下载次数:
- 60
-
高速下载
|
|
