资源描述:
《基于振动信号分形维数的变压器松动诊断方法》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、基于振动信号分形维数的变压器松动诊断方法赵莉华1,丰遥1,谢荣斌2,薛静2,张霖2,王仲1(1.四川大学电气信息学院,成都10065;2.贵州电网有限责任公司贵阳供电局,贵阳550000)摘要:基于变压器松动前后振动信号非线性特性的改变,提出采用振动信号时域波形的分形维数作为铁心、绕组松动诊断特征量。文中探讨了分形维数应用于变压器振动信号分析的可行性,根据分形理论,针对变压器振动模拟信号与实测信号分别进行了分形盒维数计算。结果表明,正常状态下不同变压器绕组振动信号分形盒维数集中于1.19,不同变压器铁心振动信号分形盒维数不同,但对于同一台变压器,其铁心振动信号分
2、形盒维数具有稳定值。变压器松动故障后,铁心和绕组振动信号的分形维数均明显增大,分形维数能准确反映变压器正常与松动状态。关键词:分形维数;振动信号;变压器;故障诊断中图分类号:TM41文献标识码:A文章编号:1001-1390(2017)00-0000-00Transformerfaultdiagnosismethodbasedonvibrationsignals’fractaldimensiodimensionofvibrationsignalsZhaoLihua1,FengYao1,XieRongbin2,XueJing2,ZhangLin2,Wang,Wan
3、gZhong1(1.SchoolofElectricalInformation,SichuanUniversity,Chengdu610065,China.2.GuiyangPowerSupplyBureauofGuizhouStateGridCo.,Ltd.,Guiyang550000,China)Abstract:Basedonthechangeofnon-linearcharacteristicsofThevibrationsignalsbeforeandaftertheoftransformersloosenessusuallyappearwiththe
4、non-linearcharacteristics.,Inviewsofthisproblem,atransformerfaultdiagnosismethodbasedonthefractaltheoryfractaldimensionofthevibrationsignalstimedomainwaveformwasproposedtoadoptasthediagnosticcharacteristicofironcoreandwindinglooseness.Thispaperdiscussedthefeasibilityoftheapplicationo
5、ffractaldimensionintheanalysisoftransformervibrationsignals.Inthispaper,kindsoftransformers’vibrationsignalsweremeasuredandthefractaldimensionofthevibrationsignalswascalculatedaccordingtofractaltheory.Theresultsshowthatthefractaldimensionofdifferentkindsofwindingsignalsoffocuson1.19u
6、ndernormalstate.Thefractaldimensionofironcorevibrationvariedwithtransformers,whileitbecamestablewiththesametransformer.Thefractaldimensionofironcoreandwindingvibrationsignalssignificantlyincreasedafterthetransformerlooseness,itcanaccuratelyreflectthenormalandloosestateoftransformer.a
7、doptedasacharacteristicparametertojudgethetransformerworkingconditions.Keywords:fractaldimension,vibrationsignal,transformer,faultdiagnosis0引言变压器作为电力系统的关键设备,其安全性是电网稳定运行的基础。铁心松动,绕组变形等机械故障目前还没有广泛应用的在线监测手段,研究学者提出了基于振动信号的变压器状态监测技术,由于和系统无电气连接,灵敏度高,安全可靠等优点,该技术在变压器铁心绕组状态的在线监测领域具有良好的应用前景。基于振
8、动信号的变压器故障诊断技