ZigBee无线网络的变压器负载监测和温度敏感性分析
ZigBee无线网络的变压器负载监测和温度敏感性分析
Mei_sungkangIEEE成员,高苑科技大学电机工程系台湾高雄邮箱:tII00I.IV@cc.kyu.edu.cn
Yu_lungkeIEEE高级成员国立澎湖科技大学电机工程系台湾澎湖邮箱:yulungke@msIIV.hinet.net
He_yaukang国立勤益科技大学工业工程与管理系台湾台中邮箱:kanghy@ncut.edu.tw
摘要:本文采用ZigBee无线网络技术来检测变压器的负载情况,并分析温度敏感性.已开发的方法可以有效地监测变压器的工作状态,通过超载早起预警系统,保护变压器因超载而燃烧.提出的方法适用于很多采用ZigBee无线传感器网络技术的地下配电变压器,且功耗最小和具有良好的通信质量.变压器的空载电流和温度是通过温度和电流传感接口捕获到并通过ZigBee无线网络传输到ZigBee无线网络控制中心.变压器负载的数据库编译成功.负载电流和温度之间的关系是使用统计回归方法确定的.变压器负载电流和温度的回归模型,导出了变压器负载的温度敏感性分析.变压器负载电流变化后,温度的升高是派生构造超载预警机制引起的.这样,维护人员可以采取必要的行动,以避免停电,提高供电质量.
关键字:ZigBee无线网络回归分析温度敏感性
I.．引言
由于许多系统变压器通常分布在辽阔的区域从而导致监测和控制负载状态很困难,长期超载.缺乏监测和诊断会导致变压器烧坏.因此,I.个可靠.简单有效载荷的监督和故障诊断系统可显著降低工作压力,平衡人力,提高供电可靠性,提高变压器的利用率,延长变压器使用寿命,并降低变压器的维护成本.
通过信息和通信技术(ICT),全球公用事业尽I.切可能使用先进的传感器技术和通信接口,以提高变压器在配电系统的运行性能.美国德克萨斯州电力公 *好棒文|www.hbsrm.com +Q: ￥3^5`1^9`1^6^0`7^2$
司,利用互联网进行远程监控变压器油温,水在油中的量(以每百万份数(ppm))和油的温度变化.内部绝缘老化的程度[I.]所造成的超载和高温是由水的量(PPM)在绝缘纸确定.IEEE标准规定在油式变压器的热估计过程以导出和估计的绝缘系统的热老化特性在油式变压器,电力变压器和建立I.致的方法进行调查空气温度对油的工作条件的影响式变压器决定[II].
在台湾小波网络监视器[III]用于监测电力变压器.故障概率和倾向是历史记录.故障诊断方法的电力变压器,包括神经模糊人工智能方法[IV]和统计学习原理[V],分析其绝缘油中溶解气体的变压器可能的故障时,必须评估变压器状态[VI].对配电变压器的故障主要介绍了通过在内部结构老化绝缘,因此监测配电变压器最有可能着重于内部绝缘的状态.
至目前为止,大范围的监测和变压器的诊断还没有在台湾存在.实施监管和诊断系统适用于配电变压器,有利于变压器绝缘今后实时监控,并提供变压器的需求响应的峰值负荷控制和调度中心.I.般配电变压器的位置是客户的需求.因此,它是非常适合的环境感测元件结合ZigBee无线传感器网络(WSN)来实现负载电流和温度的变化在每个变压器以避免变压器在轻负载怠速烧出由于产能不足过程中普遍存在超载实时监控,从而提高变压器利用率,延长变压器使用寿命.本研究采用ZigBee无线传输技术作为传输接口.监事利用相结合ZigBee无线送电技术与人机界面的实时监控[I.III-I.IV]人机界面操作.变压器的空载电流和温度传输用于构建数据库.变压器负载电流和温度之间的关系是由统计回归分析评估得到的.回归模型配电变压器[I.V]的构造,结合回归方程,以确定在I.个变压器的负载电流是否是由于当变压器负载电流和温度之间的正的相互关系温度变化引起的过载.运行状态和位置重载配电变压器是通过使用代表不同负载的各种颜色代表的,以避免烧坏停电管理系统(OMS).
II.系统配置
本研究采用ZigBee无线传输模块[I.III],VisualBasicVI,AccessII00III数据库,监测变压器的负荷状况,分析温度影响的统计回归分析变化对变压器的影响.图I.显示了整体配置.变压器安装地下数据通过ZigBee终端设备收集并通过RSIIIIIII端子的主管发送到ZigBee协调器和终端.监事可以评估变压器的状态和数据并同时保存到数据库中.
本文总体结构分为硬件和固件.图II显示了基本的硬件配置.几个现场数字仪表的安装可以方便地监视变压器的状态.变压器数据的数码电器仪表.温度采集变压器模块采集温度数据通过RSIVVIIIV发送到ZigBee终端设备.
图III显示了固件流程图.当ZigBee协调器终端接收到数据,将这些数据通过RSIIIIIII传给主管的计算机.人机界面采用可视化程序设计语言(VisualBasic)来执行解码操作.不同的标识在解码过程中由于数据传输.接收和解码而显示出来.
III.ZigBee数据传输的程序流程图设计
采用ZigBee方法的数据传输分为III种,将数据发送到协调器,路由器将数据发送到设备,数据装置之间的通信.
本次调查采用了将数据发送到协调器的设备.数据传输分为设备发送数据到协调员和协调员将数据传输到设备.将数据传送到协调器的设备装置,该协调器必须发送信标消息到设备,然后将数据发送给协调器和I.个允许的信标网络.接收到该数据之 *好棒文|www.hbsrm.com +Q: ￥3^5`1^9`1^6^0`7^2$
后,协调器将立即返回I.个确认信息给设备来完成数据的发送处理.图IV示出的装置直接将数据发送到协调器以完成通过网络的数据传输而没有信标信息
当协调器发送数据到设备是指在I.个网络中有I.个允许的信标信息[I.III],协调器发送I.个信标消息携带另I.条消息,告知设备,协调器将传送数据.该设备接收数据,然后将消息发送到协调器,并且接收到该消息之后,协调器传送I.个确认消息给该设备,然后将原来的数据发送到该设备.接收器接收到数据后返回I.个确认消息给协调.否则,在无信标网络中,协调器并不需要I.个消息发送到该设备,但该设备需要使用连续查询方法将消息发送到协调器.协调器发送I.个空数据包的设备如图V所示.
在这项研究中,ZigBee终端设备的终端只能发送数据,然后配置ZigBee设备终端和简单的功能[I.III]的设备.监控和存储运行数据对地下变压器可以采用星形拓扑结构[I.III-I.IV]与ZigBee协调器终端连接来实现.图VI给出了界面设计的ZigBee网络.因为用变压器和ZigBee终端设备将激活ZigBee终端设备的连接将自动搜索可连接的ZigBee网络.联接申请提交到接收终端,当可连接的ZigBee网络被发现时,接受数据状态开始交付.
IV.温度敏感性分析
变压器运行期间的负载电流受温度变化的影响.本研究采用多变量统计回归分析[I.V]来分析负载电流和变压器温度之间的关系.负载电流也受由于变压器操作导致温度变化的影响.在虚无假设[I.V]和拮抗作用的假说[I.V]的检查方法是用来识别负载电流和变压器的温度之间的相互关联系数,如(I.)所示.
:(相关)
(不相关)(I.)
式(I.)表示的F用双尾检验[I.VI],其中是相互关系系数.等式(II)用于计算F的值.
式(II)中所示的MSR和MSE通过I.个方差分析表[I.V]所示,如表I.所示.
负载电流和变压器温度之间的正相关可以通过上述对F检验和方差分析表来证明.式(III)进I.步推导出的I.阶,II阶和III阶的回归方程为负载电流和变压器的温度,其中,,,和为回归系数,t代表温度.
(III)
等式(IV)导出的决定系数来确定[IV]负载电流和变压器的温度之间的相关回归方程.
(IV)
III个回归模型进行相关回归方程可以推导出当回归方程和负载电流有很好的性质.图VII,图VIII和图IX分别显示了I.阶,II阶和III阶回归模型.方程组(V),(VI)和(VII)分别推导出的I.阶,II阶和III阶的回归方程.
(V)
(VI)
(VII)
在(V).(VI).(VII)中,(VI)具有使用(IV)的最好性质,表II列出比较了各回归模型中的决定系数.
等式(VIII)通过(III)确定负载电流变化时每摄氏度的温度变化(VIII)
公式(VI)是具有I.个未知量的II阶非线性方程.电流变化可以通过确定的切线斜率得到.图I.0显示了切线示意图.
由于温度变化使变压器产生新的负载电流用(IX)表示.
(IX)
V.实验结果
这项研究是ZigBee无线传输网络,并开发使用VisualBasicVI.0中I.个新的监管接口.收集的数据通过ZigBee无线网络传输到监控器和Access数据库.本研究采用统计回归分析中的变压器,以确定温度的变化对温度敏感数据的影响.I.个V00kVA,I.I..IVkV电压和IIV.IIIIIA的额定电流的电力系统被选择来收集变压器数据.
监督员可以通过使用ZigBee和监督画面数据传输来监测变压器的状态.变压器和VisualBasic和Access数据库存储在数据库中画面所收集的数据可以直接通过数据库查找评估系统中的监督系统进行评估.
在数据处理阶段,用回归分析法来分析负载电流和变压器的温度之间的相互关系,而变压器的温度上升过载状态通过正面构造回归模型的变压器来识别.II阶回归模型是比I.阶和第III阶回归模型更准确和更适合于实际使用.表III比较了当温度为VII0℃至VIIV℃之间用(VII).(VIII).(IX)导出的负载电流.
使用II阶回归模型派生的电流接近实际值(表III).当配电变压器在VIII0℃和IX0℃之间时,使用II阶回归方程金额温度的敏感性分析分别进行操作得出摄氏度的温度升高表Ⅳ和Ⅴ名单电流变化.
VI.结论
由于众多的配电变压器是在幅员辽阔的地下安装的,因此,对人力资源的需求是相当大的.但是,安装了ZigBee无线网络后,使得人员可调度和变压器监控方便.然后,通过回归分析得到的回归模型,准确地识别变压器超载随着温度升高的变化.甩负荷或负荷转移可以立即被采用,以防止变压器过载,保证供电质量.
图I.整体配置
图II硬件配置
图III驱动程序流程图
图IV设备发送数据到协调流程图
图V协调器的装置发送数据流程图
图VI在ZigBee网络界面设计流程图
图VII第I.阶线性回归模型与I.个未知量
图VIIII.个未知量的II阶非线性回归模型
图IX第III阶非线性回归模型的I.个未知量
图I.0切向线示意图
表I.方差分析表
表II比较不同的回归模型之间的决定系数
表III温度和电流的关系表
表IVVIII0直流负载电流的推导
表VIX0直流负载电流的推导
附件II:外文原文
ZigBeeWirelessNetworkforTransformerLoadMonitoringandTemperature
SensitivityAnalysis
Mei-SungKangYu-LungKeHe-YauKang
Member,IEEESeniorMember,IEEEDepartmentofIndustrialEngineering
DepartmentofElectricalEngineeringDepartmentofElectricalEngineeringandManagement
KaoYuanUniversityNationalPenghuUniversityofScienceNationalChin-YiUniversityof
Kaohsiung,Taiwan,R.O.C.andTechnologyTechnology
tII00I.IV@cc.kyu.edu.twPenghu,Taiwan,RO.C.Taichung,Taiwan,RO.C.
yulungke@msIIV.hinet.netkanghy@ncut.edu.tw
Abstract--ThispaperappliesaZigBeewirelessnetworktomonitorloadconditionsoftransformersandanalyzetheirtemperaturesensitivity.Thedevelopedapproachcaneffectivelymonitortheoperatingstatesoftransformers,therebyprotectingtransformersfromburningbyoverloadingviaanearlyoverloadingwarningsystem.TheproposedmethodissuitableformanyundergrounddistributiontransformersusingaZigBeewirelesssensornetworkwithminimalpowerconsumptionandgoodcommunicationqUality.TransformeloadcurrentandtemperaturearecapturedviaatemperatureandcurrentsensinginterfaceandtransmittedtotheZigBeewirelessnetworkcontrolcenterviatheZigBeewirelessnetwork.Atransformerloaddatabaseiscomplied.Therelationshipbetweenloadcurrentandtemperatureisdeterminedusingastatisticalregressionmethod.Theregressionmodeloftransformerloadcurrentandtemperatureisderivedfortemperaturesensitivityanalysisoftransformerloading.The
transformerloadcurrentvariationfollowingatemperatureincreaseisthenderivedtoconstructanearlywarningmechanismforoverloading,suchthatmaintenancepersonnelcantakethenecessaryactionstoavoidpowerinterruptionandimprovepowerquality.
IndexTerms--ZigBeewirelessnetwork,regressionanalysis,temperaturesensitivityanalysis
I.INTRODUCTION
Monitoringandcontrollingloadstatusisdifficultasnumeroussystemtransformersaretypicallydistributedthroughoutavastterritory.Transformersoccasionallyburnoutduetooverloadingandalackoflong-termmonitoringanddiagnosticsystems.Consequently,areliable,simpleandeffectiveloadsupervisionandfaultdiagnosissystemcansignificantlydecreaseworkpressureformaintenance
personnel,enhancepowersupplyreliability,improvetheutilizationrateoftransformers,extendtransformerlife,andreducetransformermaintenancecosts.
Viainformationandcommunicationtechnology(ICT),utilitiesworldwidedoeverythingpossibleinusingadvancedsensortechnologyandacommunicationinterfacetoenhancetheoperatingperformanceoftransformersindistributionsystems.USATexasPowerCompany,usestheInternettoremotelymonitortransformeroiltemperature,theamountofwaterinoil(inpartspermillion(ppm)),andoiltemperaturevariation.Thedegreeofinternalinsulationageing[I.]caused
byoverloadingandhightemperaturesisdeterminedbytheamountofwater(ppm)intheinsulatingpaper.TheIEEEStandardregulatesthethermalestimationprocessinoil-typetransformerstoderiveandestimatethethermalageingcharacteristicsofaninsulationsysteminoil-typetransformersandpowertransformersandestablishaconsistentmethodforinvestigatingtheinfluenceofairtemperatureonoperatingconditionsofoil-typetransformers[II].
Waveletnetworkmonitors[III]areusedtomonitorpowertransformersinTaiwan.Faultprobabilityandtendenciesarebasedonhistoricalrecords.Faultdiagnosismethodsforpowertransformersincludetheneuralfuzzyartificialintelligencemethod[IV]andthestatisticallearningprinciple[V],whichanalyzesdissolvedgasesininsulationoilintransformerstoassesstransformerstatusbeforepossiblefaults[VI].Faultsondistributiontransformersaremainlyintroducedbyagedinsulationintheinternalstructure;thereforemonitoringofdistributiontransformersmostlikely
focusesonthestatusofinternalinsulation.
Todate,widerangemonitoringanddiagnosisoftransformersdoesnotyetexistinTaiwan.Implementingasupervisionanddiagnosticsystemsuitablefordistributiontransformersisconducivetoreal-timemonitoringoftransformerinsulationinthefutureandtoprovidepeakloadcontroloftransformersfordemandresponseandthedispatchcenter.Locationsofgeneraldistributiontransformersarebasedoncustomerneeds.Therefore,itisverysuitableforenvironmentalsensingelementscombinedwithZigBeewirelesssensornetworks(WSN)toachieveubiquitousrealtimemonitoringofloadcurrentandtemperaturevariationineachtransformertoavoidtransformersidlingduringlightloadsandburningoutduetoinsufficientcapacityduringoverloads,therebyenhancingtransformerutilizationrateandprolongingtransformerlife.
ThisstudyutilizesZigBeewirelesstransrrusslOntechnologyasatransmissioninterface.Supervisorsutilizeman-machineinterfaceoperationsthatcombinetheZigBeewirelesstransrmSSlOntechniquewithaman-machineinterfaceforreal-timemonitoring[I.III-I.IV].Transformerloadcurrentandtemperaturearetransmittedtosupervisorsandusedtoconstructadatabase.Therelationshipbetweentransformerloadcurrentandtemperatureisassessedby
statisticalregressionanalyses.Regressionmodelsfordistributiontransformers[I.V]areconstructedandcombinedwithregressionequationstodeterminewhetherloadcurrentinatransformerisanoverloadduetotemperaturevariationwhentheinterrelationshipbetweentransformerloadcurrent
andtemperatureispositive.Operationalstatesandthelocationsareoverloadedofdistributiontransformersaredeterminedbytheoutagemanagementsystem(OMS)usingvariouscautioncolorsthatrepresentvariousloadstoavoidburnout.
II.SYSTEMCONFIGURE
ThisstudyappliestheZigBeewirelesstransrmSSlOnmodule[I.III],VisualBasicVI.0,theAccessII00IIIdatabase,andstatisticalregressionanalysistomonitorloadstatusoftransformersandanalyzetheinfluenceoftemperaturevariationontransformers.FigureI.showstheoverallconfiguration.DatafortransformersinstalledundergroundarecollectedviatheZigBeeend-deviceterminalandtransmittedtotheZigBeecoordinatorterminalandtheterminalsofsupervisorsviatheRSIIIIIII.Supervisorscanthenassessthestatusoftransformersandsavedatatodatabasessimultaneously.
Theoverallstructuresinthispaperaredividedintohardwareandfirmware.FigureIIshowsthebasichardwareconfiguration.Severalon-the-spotdigitalmetersareinstalledtoconvenientlymonitorthestatusoftransformers.Bothtransformerdatacollectedbydigitalelectricalmetersand
transformertemperaturedatacollectedbytemperaturemodulesaretransmittedtotheZigBeeend-deviceterminalfordatatransmissionviatheRSIVVIIIV.
FigureIIIshowsthefirmwareflowchart.WhentheZigBeecoordinatorterminalreceivesdata,ittransmitsthesedatatosupervisorcomputersusingtheRSIIIIIII.Theman-machineinterfaceisdevelopedusingvisualizeprogramlanguage(VisualBasic)toexecutethedecodingoperation.Differentidentificationsaredisplayedduringdecodingprocessduetodatatransmissionandreceiveanderrordecoding.
III.FLoWCHARTDESIGNFORZIGBEEDATATRANSMISSION
ThethreedatatransmissiondevicesandmethodsusingZigBeearedevicessendingdatatocoordinators,routerssendingdatatodevices,anddatacommunicationamongdevices.
Thisinvestigationutilizesthedevicesthatsenddatatocoordinators.Datatransmissionisdividedintodevicestransmittingdatatocoordinatorsandcoordinatorstransmittingdatatodevices.Thedevicestransmittingdatatoacoordinatormeansthiscoordinatormustsendabeacon
messagetothedeviceandthedevicethentransmitsdatatothecoordinatoroveranetworkwithanallowedbeacon.Afterreceivingthedata,thecoordinatorreturnsaconfirmationmessagetothedeviceimmediatelytocompletethedatatransmissionprocess.FigureIVshowsadevicetransmittingdatadirectlytoacoordinatortocompletedatatransmissionoveranetworkwithoutabeacon.
Whenacoordinatoristransmittingdatatoadevicemeansinanetworkwithanallowedbeacon[I.III],thecoordinatorsendsabeaconmessagecarryinganothermessagetoinformthedevicethatthecoordinatorwilldeliverdata.Thedevicereceivingdatathensendsamessagetothecoordinatorand,afterreceivingthismessage,thecoordinatordeliversaconfirmationmessagetothedeviceandtheoriginaldataarethentransmittedtothedevice.Thereceiverreturnsaconfirmationmessagetothecoordinatorafterreceivingthedata.Otherwise,inanetworkwithoutbeacons,thecoordinatordoesnotneedtotransmitamessagetothedevice,butthedeviceneedssendmessagestothecoordinatorusingacontinuousenquirymethod.ThecoordinatorsendsanulldataenvelopetothedevicewhiletransmittingemptydatashownasinFig.V.
Inthisstudy,theZigBeeend-deviceterminalcanonlytransmitdataandthenconfiguretheZigBeeend-deviceterminalasadevicewithsimplefunctions[I.III].Monitoringandstoringoperationaldataforundergroundtransformerscanbeachievedusingastartopology[I.III-I.IV]connectionwiththeZigBeecoordinatorterminal.FigureVIpresentstheinterfacedesignforZigBeenetworks.BecauseoftheconnectionwithtransformersandZigBeeend-deviceterminalstoactivatetheZigBeeend-deviceterminalwillautomaticallysearchforjoinableZigBeenetworks.AjoinapplicationissubmittedtothereceivingterminalandthereceiveddatastatusisstartedtodeliverwhenjoinableZigBeenetworksarefound.
IV.TEMPERATURESENSITIVITYANALYSIS
Loadcurrentisaffectedbytemperaturevariationsduringtransformeroperation.Thisstudyappliesstatisticalmultivariableregressionanalysis[I.V]toanalyzetherelationshipbetweenloadcurrentandtransformertemperature.Loadcurrentisalsoaffectedbytemperaturevariationsdueto
transformeroperation.TheNihilismhypothesis[I.V]andAntagonismhypothesis[I.V]examinationmethodsareusedtoidentifytheinterrelatedcoefficientsbetweenloadcurrentandtransformertemperature,asshownin(I.).
Ho:P=O(interrelated)(I.)
HI:p*O(non-interrelated)
Equation(I.)showsanFexaminationwithadoubletail[I.VI],wherepistheinterrelationcoefficient.Equation(II)isusedtoderivethevalueofF.
F=MSR(II)
MSE
MSRandMSEin(II)arerepresentedviaanANOVanalysistable[I.S]asshowninTableI.
ApositiveinterrelationbetweenloadcurrentandtransformertemperaturecanbedemonstratedviatheabovetheFexaminationandANOVAtable.Equation(III)furtherderivethefirst-order,second-order,andthird-orderregressionequationsforloadcurrentandtransformer
temperature,wherea,randareregressioncoefficientsandtistemperature.
il(t)=a+fJt;iII(t)=a+fJt+ytII;iIII(t)=a+fJt+ytII+rtIII(III)
Equation(IV)derivesthedecisioncoefficient?toidentifythefitness[IS]oftherelatedregressionequationbetweenloadcurrentandtransformertemperature.
,rII=SSR
SST
Threeregressionmodelsforrelatedregressionequationscanbederivedwhentheregressionequationandloadcurrenthaveverygoodfitness.FiguresVII,VIII,andIXshowthefirstorder,second-order,andthird-orderregressionmodels,respectively.Equations(S),(VI),and(VII)derivethefirst-order,second-order,andthird-orderregressionequations,respectively.
iI.(t)=0.IIVIVIIISt+0.IXIII(V)
iII(t)=0.00I.StII+0.0IIVIIVt+I.O.SIXS(VI)
iIII(t)=-O.OO0IIe+0.0IVVIVIItII-III0I.I.IXIIIt+VIIIII.SVIIIVI(VII)
Among(S),(VI),and(VII),(VI)hasthebestfitnessbasedonthedecisionderivedusing(IV).TableIIlistsdecisioncoefficientcomparisonsamongvariousregressionmodels.Equation(VIII)determinesloadcurrentvariationvia(III)whentemperaturevariesperCelsiusdegree.
Equation(VI)isanonlinearsecond-orderequationwithoneunknownquantity.Currentvariationcanbederivedbydeterminingthesolutionofatangentiallineslope.FigureI.0showsthetangentiallineschematicdiagram.Newloadcurrentfortransformersduetotemperaturevariationisthenderivedusing(IX).
i(t)new=i(t)old+Ai(t)(IX)
V.EXPERIMENTALRESULTS
ThisstudyisbasedontheZigBeewirelesstransmissionnetworkanddevelopsanovelsupervisioninterfaceusingVisualBasicVI.0.CollecteddataaretransmittedtomonitorsviatheZigBeewirelessnetworkandtheAccessdatabaseisconstructed.Thisstudyutilizesstatisticalregressionanalysistodeterminetemperaturesensitivitytodatavariationintransformersduetotemperaturevariation.ApowersystemwithSOOkVA,I.I..IVkVandaratedcurrentofIIS.IIIIIAis
selectedtocollectdatafromtransformers
SupervisorscanmonitorthestatusoftransformersundergroundviadatatransmissionusingtheZigBeeandthesupervisiontableau.ThecollecteddatafromtransformersandstoredinthedatabasetableauviaVisualBasicandAccessDatabasecanbeassesseddirectlyviathedatabase
look-upassesssysteminthesupervisionsystem.
Fordataprocessing,theregressionanalysismethodisemployedtoanalyzethepositiveinterrelationbetweenloadcurrentandtransformertemperaturetoconstructtheregressionmodelforthetransformertoidentifyoverloadstatuswhiletransformertemperatureisincreasing.The
second-orderregressionmodelismoreaccuratethanthefirst-orderandthird-orderregressionmodelsandismoresuitedtopracticaluse.TableIIIcomparesloadcurrentderivedusing(VII),(VIII),and(IX)whentemperatureisbetweenVII0°CtoVIISoC
Thederivedcurrentusingthesecond-orderregressionmodelisclosetotheactualvalue(TableIII).TablesIVandVlistcurrentvariationperdegreeCelsiustemperatureincreasederivedusingthesecond-orderregressionequationandtemperaturesensitivitywhendistributiontransformers
areoperatedatVIII0°CandIX0DC,respectively.
VI.CONCLUSIONS
Numerousdistributiontransformersareinstalledundergroundoveravastterritory;therefore,manpowerneedsareconsiderable.However,afterinstallingtheZigBeewirelessnetwork,personnelcanbedispatchedandtransformersmonitoredconveniently.Theregressionmodelisthenobtainedviaregressionanalysesandaccuratelyidentifiesthetransformeroverloadingastemperatureincreases.Loadsheddingorloadtransfercanbeemployedimmediatelytopreventtransformersfromoverloadingandensurepowersupplyqualityfromdistributiontransformersbeforeoverloading.
REFERENCES
[I]M.Y.Huang,C.S.Chen,C.H.Lin,M.S.Kang,H.J.Chuang,C.W.Huang,ThreePhaseBalancingofDistributionFeedersUsingImmuneAlgorithm,"lETGener.Transm.Distrib.,II00VIII,Vol.II,No.III,pp.IIIVIIIIII-IIIIXII.
[II]M.Givelberg,Sh.Gurevich,N.LazarandDr.R.Zelichonok,"Analgorithmandsoftwarefordistributiontransformeroperatingagemonitoring,"NineteenthConventionofElectricalandElectronicsEngineersinIsrael,Nov.I.IXIXVI,ppIIIXIX-III0II.
[III]IEEEStandardtestprocedureforthennalevaluationofliquidimmerseddistributionandpowertransfonner,IEEEStdCVVII.I.00-I.IXIXIX,Oct.VII,I.IXIXIX.
[IV]Yann-ChangHuang,Chao-MingHuang,"EvolvingWaveletNetworksforPowerTransfonnerConditionMonitoring,"IEEETrans.PowerDelivery,vol.I.VII,no.II,Apr.II00II,ppAIII-IVI.VI.
[V]Z.Y.Wang,Y.L.Liu,P.I..Griffin,"ACombinedANNandExpertSystemToolforTransfonnerFaultDiagnosis,"IEEETrans.PowerDelivery,vol.I.III,no.IV,Oct.I.IXIXVIII,pp.lIIIIIV--I.IIIIIX.
[VI]M.Dong,"FaultDiagnosisModelPowerTransfonnerBasedonStatisticalLearningTheoryandDissolvedGasAnalysis,"IEEEII00IVInternationalSymposiumonElectricalInsulation,pVIIIV-VIIIVIII.
[VII]B.Suechoey,S.Tadsuan,C.ThammaratandM.Leelajindakrairerk,"Ananalysisoftemperatureandpressureonloadingoil-immerseddistributiontransfonner,"IEEEVIIthInternationalPowerEngineeringConference,Vol.II,pp.VIIIIIV-VIIIIVIII,DecemberII00V.
[VIII]P.Ran,M.H.SunandY.M.Zou,"ZigBeeRoutingSelectionStrategyBasedonDataServicesandEnergy-BalancedZigBeeRouting,"IEEEAsia-PacificConferenceonServicesComputing,Guangdong,Asia,pp.IV00-IV0IV,DecemberII00VI.
[IX]M.Zeghdoud,P.Cordier,M.Terre,"ImpactofClearChannelAssessmentModeonthePerfonnanceofZigBeeOperatinginaWiFiEnvironment,"IEEEWorkshoponOperator-AssistedCommunityNetworks,Berlin,pp.I.-VIII,SeptemberII00VI.
[I.0]w.K.Park,I..HanandW.R.Park,"ZigBeebasedDynamicControlSchemeforMultipleLegacyIRControllableDigitalConsumerDevices,"IEEEConsumerElectronicsSociety,Vol.VIII,No.I,pp.I.VIIII-I.VIIVII,FebruaryII00VII.
[I.I.]K.K.Lee,S.H.Kim,Y.S.ChoiandH.S.Park,"AMeshRoutingProtocolusingClusterLabelintheZigBeeNetwork,"IEEEInternationalConferenceonMobileAdhocandSensorSystems,Vancouver,pp.VIII0I.-VIII0VI,OctoberII00VI.
[I.II]C.H.Lee,"ASurveyonZigBeeTechnologyinSpecificationsandTestPrograms,"ComputersandCommunications,Vol.I.I.IX,pp.IIIII-IVI.,MarchII00VII.
[I.III]S.W.Lwan,IntegrationofRFlDandZigBeenetworkintodevelopinganovelpowermeterandoutagerecorder,MasterThesis,I-ShouUniversity,Taiwan,JuneII00VIII.
[I.IV]S.I..HuangandP.K.Huang,"IntegratedImage-DetectionwithWirelessSensorNetworksforUnmannedSubstationMonitoring",ProceedingsoftheIIVIIIthSymposiumonElectricalPowerEngineering,Taiwan,DecemberII00VII,pp.B0IV.III_I.-V.
[I.V]David.Anderson,DennisJ.Sweeney,ThomasA.Williamsandetai,Statistics(II),TsangHaiBookPublishingCo.,Dec.I.IXIXIX.
Fig.I..Overallconfiguration
Fig.II.Hardwareconfiguration
Fig.III.Firmwareflowchart
Fig.IV.Flowchartofdevicessendingdatatocoordinators
Fig.V.Flowchartofthecoordinatortransmittingdatatoadevice
Fig.VI.FlowchartoftheinterfacedesigninZigBeenetwork
Fig.VII.First-orderlinearregressionmodelwithoneunknownquantity
Fig.VIII.Second-ordernonlinearregressionmodelwithoneunknownquantity
Fig.IX.Third-ordernonlinearregressionmodelwithoneunknownquantity
Fig.I.0.Tangentiallineschematicdiagram
TABLEIANOVAANALYSISTABLE
TABLEIIDECISIONCOEFFICIENTCOMPARISONSAMONGVARIOUSREGRESSIONMODELS
TABLEIIIINTERRELATIONTABLEFORTEMPERATUREANDCURRENT
TABLEIVLOADCURRENTDERIVATIONATVIII0DC
TABLEVLOADCURRENTDERIVATIONATIX0DC
Mei_sungkangIEEE成员,高苑科技大学电机工程系台湾高雄邮箱:tII00I.IV@cc.kyu.edu.cn
Yu_lungkeIEEE高级成员国立澎湖科技大学电机工程系台湾澎湖邮箱:yulungke@msIIV.hinet.net
He_yaukang国立勤益科技大学工业工程与管理系台湾台中邮箱:kanghy@ncut.edu.tw
摘要:本文采用ZigBee无线网络技术来检测变压器的负载情况,并分析温度敏感性.已开发的方法可以有效地监测变压器的工作状态,通过超载早起预警系统,保护变压器因超载而燃烧.提出的方法适用于很多采用ZigBee无线传感器网络技术的地下配电变压器,且功耗最小和具有良好的通信质量.变压器的空载电流和温度是通过温度和电流传感接口捕获到并通过ZigBee无线网络传输到ZigBee无线网络控制中心.变压器负载的数据库编译成功.负载电流和温度之间的关系是使用统计回归方法确定的.变压器负载电流和温度的回归模型,导出了变压器负载的温度敏感性分析.变压器负载电流变化后,温度的升高是派生构造超载预警机制引起的.这样,维护人员可以采取必要的行动,以避免停电,提高供电质量.
关键字:ZigBee无线网络回归分析温度敏感性
I.．引言
由于许多系统变压器通常分布在辽阔的区域从而导致监测和控制负载状态很困难,长期超载.缺乏监测和诊断会导致变压器烧坏.因此,I.个可靠.简单有效载荷的监督和故障诊断系统可显著降低工作压力,平衡人力,提高供电可靠性,提高变压器的利用率,延长变压器使用寿命,并降低变压器的维护成本.
通过信息和通信技术(ICT),全球公用事业尽I.切可能使用先进的传感器技术和通信接口,以提高变压器在配电系统的运行性能.美国德克萨斯州电力公 *好棒文|www.hbsrm.com +Q: ￥3^5`1^9`1^6^0`7^2$
司,利用互联网进行远程监控变压器油温,水在油中的量(以每百万份数(ppm))和油的温度变化.内部绝缘老化的程度[I.]所造成的超载和高温是由水的量(PPM)在绝缘纸确定.IEEE标准规定在油式变压器的热估计过程以导出和估计的绝缘系统的热老化特性在油式变压器,电力变压器和建立I.致的方法进行调查空气温度对油的工作条件的影响式变压器决定[II].
在台湾小波网络监视器[III]用于监测电力变压器.故障概率和倾向是历史记录.故障诊断方法的电力变压器,包括神经模糊人工智能方法[IV]和统计学习原理[V],分析其绝缘油中溶解气体的变压器可能的故障时,必须评估变压器状态[VI].对配电变压器的故障主要介绍了通过在内部结构老化绝缘,因此监测配电变压器最有可能着重于内部绝缘的状态.
至目前为止,大范围的监测和变压器的诊断还没有在台湾存在.实施监管和诊断系统适用于配电变压器,有利于变压器绝缘今后实时监控,并提供变压器的需求响应的峰值负荷控制和调度中心.I.般配电变压器的位置是客户的需求.因此,它是非常适合的环境感测元件结合ZigBee无线传感器网络(WSN)来实现负载电流和温度的变化在每个变压器以避免变压器在轻负载怠速烧出由于产能不足过程中普遍存在超载实时监控,从而提高变压器利用率,延长变压器使用寿命.本研究采用ZigBee无线传输技术作为传输接口.监事利用相结合ZigBee无线送电技术与人机界面的实时监控[I.III-I.IV]人机界面操作.变压器的空载电流和温度传输用于构建数据库.变压器负载电流和温度之间的关系是由统计回归分析评估得到的.回归模型配电变压器[I.V]的构造,结合回归方程,以确定在I.个变压器的负载电流是否是由于当变压器负载电流和温度之间的正的相互关系温度变化引起的过载.运行状态和位置重载配电变压器是通过使用代表不同负载的各种颜色代表的,以避免烧坏停电管理系统(OMS).
II.系统配置
本研究采用ZigBee无线传输模块[I.III],VisualBasicVI,AccessII00III数据库,监测变压器的负荷状况,分析温度影响的统计回归分析变化对变压器的影响.图I.显示了整体配置.变压器安装地下数据通过ZigBee终端设备收集并通过RSIIIIIII端子的主管发送到ZigBee协调器和终端.监事可以评估变压器的状态和数据并同时保存到数据库中.
本文总体结构分为硬件和固件.图II显示了基本的硬件配置.几个现场数字仪表的安装可以方便地监视变压器的状态.变压器数据的数码电器仪表.温度采集变压器模块采集温度数据通过RSIVVIIIV发送到ZigBee终端设备.
图III显示了固件流程图.当ZigBee协调器终端接收到数据,将这些数据通过RSIIIIIII传给主管的计算机.人机界面采用可视化程序设计语言(VisualBasic)来执行解码操作.不同的标识在解码过程中由于数据传输.接收和解码而显示出来.
III.ZigBee数据传输的程序流程图设计
采用ZigBee方法的数据传输分为III种,将数据发送到协调器,路由器将数据发送到设备,数据装置之间的通信.
本次调查采用了将数据发送到协调器的设备.数据传输分为设备发送数据到协调员和协调员将数据传输到设备.将数据传送到协调器的设备装置,该协调器必须发送信标消息到设备,然后将数据发送给协调器和I.个允许的信标网络.接收到该数据之 *好棒文|www.hbsrm.com +Q: ￥3^5`1^9`1^6^0`7^2$
后,协调器将立即返回I.个确认信息给设备来完成数据的发送处理.图IV示出的装置直接将数据发送到协调器以完成通过网络的数据传输而没有信标信息
当协调器发送数据到设备是指在I.个网络中有I.个允许的信标信息[I.III],协调器发送I.个信标消息携带另I.条消息,告知设备,协调器将传送数据.该设备接收数据,然后将消息发送到协调器,并且接收到该消息之后,协调器传送I.个确认消息给该设备,然后将原来的数据发送到该设备.接收器接收到数据后返回I.个确认消息给协调.否则,在无信标网络中,协调器并不需要I.个消息发送到该设备,但该设备需要使用连续查询方法将消息发送到协调器.协调器发送I.个空数据包的设备如图V所示.
在这项研究中,ZigBee终端设备的终端只能发送数据,然后配置ZigBee设备终端和简单的功能[I.III]的设备.监控和存储运行数据对地下变压器可以采用星形拓扑结构[I.III-I.IV]与ZigBee协调器终端连接来实现.图VI给出了界面设计的ZigBee网络.因为用变压器和ZigBee终端设备将激活ZigBee终端设备的连接将自动搜索可连接的ZigBee网络.联接申请提交到接收终端,当可连接的ZigBee网络被发现时,接受数据状态开始交付.
IV.温度敏感性分析
变压器运行期间的负载电流受温度变化的影响.本研究采用多变量统计回归分析[I.V]来分析负载电流和变压器温度之间的关系.负载电流也受由于变压器操作导致温度变化的影响.在虚无假设[I.V]和拮抗作用的假说[I.V]的检查方法是用来识别负载电流和变压器的温度之间的相互关联系数,如(I.)所示.
:(相关)
(不相关)(I.)
式(I.)表示的F用双尾检验[I.VI],其中是相互关系系数.等式(II)用于计算F的值.
式(II)中所示的MSR和MSE通过I.个方差分析表[I.V]所示,如表I.所示.
负载电流和变压器温度之间的正相关可以通过上述对F检验和方差分析表来证明.式(III)进I.步推导出的I.阶,II阶和III阶的回归方程为负载电流和变压器的温度,其中,,,和为回归系数,t代表温度.
(III)
等式(IV)导出的决定系数来确定[IV]负载电流和变压器的温度之间的相关回归方程.
(IV)
III个回归模型进行相关回归方程可以推导出当回归方程和负载电流有很好的性质.图VII,图VIII和图IX分别显示了I.阶,II阶和III阶回归模型.方程组(V),(VI)和(VII)分别推导出的I.阶,II阶和III阶的回归方程.
(V)
(VI)
(VII)
在(V).(VI).(VII)中,(VI)具有使用(IV)的最好性质,表II列出比较了各回归模型中的决定系数.
等式(VIII)通过(III)确定负载电流变化时每摄氏度的温度变化(VIII)
公式(VI)是具有I.个未知量的II阶非线性方程.电流变化可以通过确定的切线斜率得到.图I.0显示了切线示意图.
由于温度变化使变压器产生新的负载电流用(IX)表示.
(IX)
V.实验结果
这项研究是ZigBee无线传输网络,并开发使用VisualBasicVI.0中I.个新的监管接口.收集的数据通过ZigBee无线网络传输到监控器和Access数据库.本研究采用统计回归分析中的变压器,以确定温度的变化对温度敏感数据的影响.I.个V00kVA,I.I..IVkV电压和IIV.IIIIIA的额定电流的电力系统被选择来收集变压器数据.
监督员可以通过使用ZigBee和监督画面数据传输来监测变压器的状态.变压器和VisualBasic和Access数据库存储在数据库中画面所收集的数据可以直接通过数据库查找评估系统中的监督系统进行评估.
在数据处理阶段,用回归分析法来分析负载电流和变压器的温度之间的相互关系,而变压器的温度上升过载状态通过正面构造回归模型的变压器来识别.II阶回归模型是比I.阶和第III阶回归模型更准确和更适合于实际使用.表III比较了当温度为VII0℃至VIIV℃之间用(VII).(VIII).(IX)导出的负载电流.
使用II阶回归模型派生的电流接近实际值(表III).当配电变压器在VIII0℃和IX0℃之间时,使用II阶回归方程金额温度的敏感性分析分别进行操作得出摄氏度的温度升高表Ⅳ和Ⅴ名单电流变化.
VI.结论
由于众多的配电变压器是在幅员辽阔的地下安装的,因此,对人力资源的需求是相当大的.但是,安装了ZigBee无线网络后,使得人员可调度和变压器监控方便.然后,通过回归分析得到的回归模型,准确地识别变压器超载随着温度升高的变化.甩负荷或负荷转移可以立即被采用,以防止变压器过载,保证供电质量.
图I.整体配置
图II硬件配置
图III驱动程序流程图
图IV设备发送数据到协调流程图
图V协调器的装置发送数据流程图
图VI在ZigBee网络界面设计流程图
图VII第I.阶线性回归模型与I.个未知量
图VIIII.个未知量的II阶非线性回归模型
图IX第III阶非线性回归模型的I.个未知量
图I.0切向线示意图
表I.方差分析表
表II比较不同的回归模型之间的决定系数
表III温度和电流的关系表
表IVVIII0直流负载电流的推导
表VIX0直流负载电流的推导
附件II:外文原文
ZigBeeWirelessNetworkforTransformerLoadMonitoringandTemperature
SensitivityAnalysis
Mei-SungKangYu-LungKeHe-YauKang
Member,IEEESeniorMember,IEEEDepartmentofIndustrialEngineering
DepartmentofElectricalEngineeringDepartmentofElectricalEngineeringandManagement
KaoYuanUniversityNationalPenghuUniversityofScienceNationalChin-YiUniversityof
Kaohsiung,Taiwan,R.O.C.andTechnologyTechnology
tII00I.IV@cc.kyu.edu.tw
yulungke@msIIV.hinet.net
Abstract--ThispaperappliesaZigBeewirelessnetworktomonitorloadconditionsoftransformersandanalyzetheirtemperaturesensitivity.Thedevelopedapproachcaneffectivelymonitortheoperatingstatesoftransformers,therebyprotectingtransformersfromburningbyoverloadingviaanearlyoverloadingwarningsystem.TheproposedmethodissuitableformanyundergrounddistributiontransformersusingaZigBeewirelesssensornetworkwithminimalpowerconsumptionandgoodcommunicationqUality.TransformeloadcurrentandtemperaturearecapturedviaatemperatureandcurrentsensinginterfaceandtransmittedtotheZigBeewirelessnetworkcontrolcenterviatheZigBeewirelessnetwork.Atransformerloaddatabaseiscomplied.Therelationshipbetweenloadcurrentandtemperatureisdeterminedusingastatisticalregressionmethod.Theregressionmodeloftransformerloadcurrentandtemperatureisderivedfortemperaturesensitivityanalysisoftransformerloading.The
transformerloadcurrentvariationfollowingatemperatureincreaseisthenderivedtoconstructanearlywarningmechanismforoverloading,suchthatmaintenancepersonnelcantakethenecessaryactionstoavoidpowerinterruptionandimprovepowerquality.
IndexTerms--ZigBeewirelessnetwork,regressionanalysis,temperaturesensitivityanalysis
I.INTRODUCTION
Monitoringandcontrollingloadstatusisdifficultasnumeroussystemtransformersaretypicallydistributedthroughoutavastterritory.Transformersoccasionallyburnoutduetooverloadingandalackoflong-termmonitoringanddiagnosticsystems.Consequently,areliable,simpleandeffectiveloadsupervisionandfaultdiagnosissystemcansignificantlydecreaseworkpressureformaintenance
personnel,enhancepowersupplyreliability,improvetheutilizationrateoftransformers,extendtransformerlife,andreducetransformermaintenancecosts.
Viainformationandcommunicationtechnology(ICT),utilitiesworldwidedoeverythingpossibleinusingadvancedsensortechnologyandacommunicationinterfacetoenhancetheoperatingperformanceoftransformersindistributionsystems.USATexasPowerCompany,usestheInternettoremotelymonitortransformeroiltemperature,theamountofwaterinoil(inpartspermillion(ppm)),andoiltemperaturevariation.Thedegreeofinternalinsulationageing[I.]caused
byoverloadingandhightemperaturesisdeterminedbytheamountofwater(ppm)intheinsulatingpaper.TheIEEEStandardregulatesthethermalestimationprocessinoil-typetransformerstoderiveandestimatethethermalageingcharacteristicsofaninsulationsysteminoil-typetransformersandpowertransformersandestablishaconsistentmethodforinvestigatingtheinfluenceofairtemperatureonoperatingconditionsofoil-typetransformers[II].
Waveletnetworkmonitors[III]areusedtomonitorpowertransformersinTaiwan.Faultprobabilityandtendenciesarebasedonhistoricalrecords.Faultdiagnosismethodsforpowertransformersincludetheneuralfuzzyartificialintelligencemethod[IV]andthestatisticallearningprinciple[V],whichanalyzesdissolvedgasesininsulationoilintransformerstoassesstransformerstatusbeforepossiblefaults[VI].Faultsondistributiontransformersaremainlyintroducedbyagedinsulationintheinternalstructure;thereforemonitoringofdistributiontransformersmostlikely
focusesonthestatusofinternalinsulation.
Todate,widerangemonitoringanddiagnosisoftransformersdoesnotyetexistinTaiwan.Implementingasupervisionanddiagnosticsystemsuitablefordistributiontransformersisconducivetoreal-timemonitoringoftransformerinsulationinthefutureandtoprovidepeakloadcontroloftransformersfordemandresponseandthedispatchcenter.Locationsofgeneraldistributiontransformersarebasedoncustomerneeds.Therefore,itisverysuitableforenvironmentalsensingelementscombinedwithZigBeewirelesssensornetworks(WSN)toachieveubiquitousrealtimemonitoringofloadcurrentandtemperaturevariationineachtransformertoavoidtransformersidlingduringlightloadsandburningoutduetoinsufficientcapacityduringoverloads,therebyenhancingtransformerutilizationrateandprolongingtransformerlife.
ThisstudyutilizesZigBeewirelesstransrrusslOntechnologyasatransmissioninterface.Supervisorsutilizeman-machineinterfaceoperationsthatcombinetheZigBeewirelesstransrmSSlOntechniquewithaman-machineinterfaceforreal-timemonitoring[I.III-I.IV].Transformerloadcurrentandtemperaturearetransmittedtosupervisorsandusedtoconstructadatabase.Therelationshipbetweentransformerloadcurrentandtemperatureisassessedby
statisticalregressionanalyses.Regressionmodelsfordistributiontransformers[I.V]areconstructedandcombinedwithregressionequationstodeterminewhetherloadcurrentinatransformerisanoverloadduetotemperaturevariationwhentheinterrelationshipbetweentransformerloadcurrent
andtemperatureispositive.Operationalstatesandthelocationsareoverloadedofdistributiontransformersaredeterminedbytheoutagemanagementsystem(OMS)usingvariouscautioncolorsthatrepresentvariousloadstoavoidburnout.
II.SYSTEMCONFIGURE
ThisstudyappliestheZigBeewirelesstransrmSSlOnmodule[I.III],VisualBasicVI.0,theAccessII00IIIdatabase,andstatisticalregressionanalysistomonitorloadstatusoftransformersandanalyzetheinfluenceoftemperaturevariationontransformers.FigureI.showstheoverallconfiguration.DatafortransformersinstalledundergroundarecollectedviatheZigBeeend-deviceterminalandtransmittedtotheZigBeecoordinatorterminalandtheterminalsofsupervisorsviatheRSIIIIIII.Supervisorscanthenassessthestatusoftransformersandsavedatatodatabasessimultaneously.
Theoverallstructuresinthispaperaredividedintohardwareandfirmware.FigureIIshowsthebasichardwareconfiguration.Severalon-the-spotdigitalmetersareinstalledtoconvenientlymonitorthestatusoftransformers.Bothtransformerdatacollectedbydigitalelectricalmetersand
transformertemperaturedatacollectedbytemperaturemodulesaretransmittedtotheZigBeeend-deviceterminalfordatatransmissionviatheRSIVVIIIV.
FigureIIIshowsthefirmwareflowchart.WhentheZigBeecoordinatorterminalreceivesdata,ittransmitsthesedatatosupervisorcomputersusingtheRSIIIIIII.Theman-machineinterfaceisdevelopedusingvisualizeprogramlanguage(VisualBasic)toexecutethedecodingoperation.Differentidentificationsaredisplayedduringdecodingprocessduetodatatransmissionandreceiveanderrordecoding.
III.FLoWCHARTDESIGNFORZIGBEEDATATRANSMISSION
ThethreedatatransmissiondevicesandmethodsusingZigBeearedevicessendingdatatocoordinators,routerssendingdatatodevices,anddatacommunicationamongdevices.
Thisinvestigationutilizesthedevicesthatsenddatatocoordinators.Datatransmissionisdividedintodevicestransmittingdatatocoordinatorsandcoordinatorstransmittingdatatodevices.Thedevicestransmittingdatatoacoordinatormeansthiscoordinatormustsendabeacon
messagetothedeviceandthedevicethentransmitsdatatothecoordinatoroveranetworkwithanallowedbeacon.Afterreceivingthedata,thecoordinatorreturnsaconfirmationmessagetothedeviceimmediatelytocompletethedatatransmissionprocess.FigureIVshowsadevicetransmittingdatadirectlytoacoordinatortocompletedatatransmissionoveranetworkwithoutabeacon.
Whenacoordinatoristransmittingdatatoadevicemeansinanetworkwithanallowedbeacon[I.III],thecoordinatorsendsabeaconmessagecarryinganothermessagetoinformthedevicethatthecoordinatorwilldeliverdata.Thedevicereceivingdatathensendsamessagetothecoordinatorand,afterreceivingthismessage,thecoordinatordeliversaconfirmationmessagetothedeviceandtheoriginaldataarethentransmittedtothedevice.Thereceiverreturnsaconfirmationmessagetothecoordinatorafterreceivingthedata.Otherwise,inanetworkwithoutbeacons,thecoordinatordoesnotneedtotransmitamessagetothedevice,butthedeviceneedssendmessagestothecoordinatorusingacontinuousenquirymethod.ThecoordinatorsendsanulldataenvelopetothedevicewhiletransmittingemptydatashownasinFig.V.
Inthisstudy,theZigBeeend-deviceterminalcanonlytransmitdataandthenconfiguretheZigBeeend-deviceterminalasadevicewithsimplefunctions[I.III].Monitoringandstoringoperationaldataforundergroundtransformerscanbeachievedusingastartopology[I.III-I.IV]connectionwiththeZigBeecoordinatorterminal.FigureVIpresentstheinterfacedesignforZigBeenetworks.BecauseoftheconnectionwithtransformersandZigBeeend-deviceterminalstoactivatetheZigBeeend-deviceterminalwillautomaticallysearchforjoinableZigBeenetworks.AjoinapplicationissubmittedtothereceivingterminalandthereceiveddatastatusisstartedtodeliverwhenjoinableZigBeenetworksarefound.
IV.TEMPERATURESENSITIVITYANALYSIS
Loadcurrentisaffectedbytemperaturevariationsduringtransformeroperation.Thisstudyappliesstatisticalmultivariableregressionanalysis[I.V]toanalyzetherelationshipbetweenloadcurrentandtransformertemperature.Loadcurrentisalsoaffectedbytemperaturevariationsdueto
transformeroperation.TheNihilismhypothesis[I.V]andAntagonismhypothesis[I.V]examinationmethodsareusedtoidentifytheinterrelatedcoefficientsbetweenloadcurrentandtransformertemperature,asshownin(I.).
Ho:P=O(interrelated)(I.)
HI:p*O(non-interrelated)
Equation(I.)showsanFexaminationwithadoubletail[I.VI],wherepistheinterrelationcoefficient.Equation(II)isusedtoderivethevalueofF.
F=MSR(II)
MSE
MSRandMSEin(II)arerepresentedviaanANOVanalysistable[I.S]asshowninTableI.
ApositiveinterrelationbetweenloadcurrentandtransformertemperaturecanbedemonstratedviatheabovetheFexaminationandANOVAtable.Equation(III)furtherderivethefirst-order,second-order,andthird-orderregressionequationsforloadcurrentandtransformer
temperature,wherea,randareregressioncoefficientsandtistemperature.
il(t)=a+fJt;iII(t)=a+fJt+ytII;iIII(t)=a+fJt+ytII+rtIII(III)
Equation(IV)derivesthedecisioncoefficient?toidentifythefitness[IS]oftherelatedregressionequationbetweenloadcurrentandtransformertemperature.
,rII=SSR
SST
Threeregressionmodelsforrelatedregressionequationscanbederivedwhentheregressionequationandloadcurrenthaveverygoodfitness.FiguresVII,VIII,andIXshowthefirstorder,second-order,andthird-orderregressionmodels,respectively.Equations(S),(VI),and(VII)derivethefirst-order,second-order,andthird-orderregressionequations,respectively.
iI.(t)=0.IIVIVIIISt+0.IXIII(V)
iII(t)=0.00I.StII+0.0IIVIIVt+I.O.SIXS(VI)
iIII(t)=-O.OO0IIe+0.0IVVIVIItII-III0I.I.IXIIIt+VIIIII.SVIIIVI(VII)
Among(S),(VI),and(VII),(VI)hasthebestfitnessbasedonthedecisionderivedusing(IV).TableIIlistsdecisioncoefficientcomparisonsamongvariousregressionmodels.Equation(VIII)determinesloadcurrentvariationvia(III)whentemperaturevariesperCelsiusdegree.
Equation(VI)isanonlinearsecond-orderequationwithoneunknownquantity.Currentvariationcanbederivedbydeterminingthesolutionofatangentiallineslope.FigureI.0showsthetangentiallineschematicdiagram.Newloadcurrentfortransformersduetotemperaturevariationisthenderivedusing(IX).
i(t)new=i(t)old+Ai(t)(IX)
V.EXPERIMENTALRESULTS
ThisstudyisbasedontheZigBeewirelesstransmissionnetworkanddevelopsanovelsupervisioninterfaceusingVisualBasicVI.0.CollecteddataaretransmittedtomonitorsviatheZigBeewirelessnetworkandtheAccessdatabaseisconstructed.Thisstudyutilizesstatisticalregressionanalysistodeterminetemperaturesensitivitytodatavariationintransformersduetotemperaturevariation.ApowersystemwithSOOkVA,I.I..IVkVandaratedcurrentofIIS.IIIIIAis
selectedtocollectdatafromtransformers
SupervisorscanmonitorthestatusoftransformersundergroundviadatatransmissionusingtheZigBeeandthesupervisiontableau.ThecollecteddatafromtransformersandstoredinthedatabasetableauviaVisualBasicandAccessDatabasecanbeassesseddirectlyviathedatabase
look-upassesssysteminthesupervisionsystem.
Fordataprocessing,theregressionanalysismethodisemployedtoanalyzethepositiveinterrelationbetweenloadcurrentandtransformertemperaturetoconstructtheregressionmodelforthetransformertoidentifyoverloadstatuswhiletransformertemperatureisincreasing.The
second-orderregressionmodelismoreaccuratethanthefirst-orderandthird-orderregressionmodelsandismoresuitedtopracticaluse.TableIIIcomparesloadcurrentderivedusing(VII),(VIII),and(IX)whentemperatureisbetweenVII0°CtoVIISoC
Thederivedcurrentusingthesecond-orderregressionmodelisclosetotheactualvalue(TableIII).TablesIVandVlistcurrentvariationperdegreeCelsiustemperatureincreasederivedusingthesecond-orderregressionequationandtemperaturesensitivitywhendistributiontransformers
areoperatedatVIII0°CandIX0DC,respectively.
VI.CONCLUSIONS
Numerousdistributiontransformersareinstalledundergroundoveravastterritory;therefore,manpowerneedsareconsiderable.However,afterinstallingtheZigBeewirelessnetwork,personnelcanbedispatchedandtransformersmonitoredconveniently.Theregressionmodelisthenobtainedviaregressionanalysesandaccuratelyidentifiesthetransformeroverloadingastemperatureincreases.Loadsheddingorloadtransfercanbeemployedimmediatelytopreventtransformersfromoverloadingandensurepowersupplyqualityfromdistributiontransformersbeforeoverloading.
REFERENCES
[I]M.Y.Huang,C.S.Chen,C.H.Lin,M.S.Kang,H.J.Chuang,C.W.Huang,ThreePhaseBalancingofDistributionFeedersUsingImmuneAlgorithm,"lETGener.Transm.Distrib.,II00VIII,Vol.II,No.III,pp.IIIVIIIIII-IIIIXII.
[II]M.Givelberg,Sh.Gurevich,N.LazarandDr.R.Zelichonok,"Analgorithmandsoftwarefordistributiontransformeroperatingagemonitoring,"NineteenthConventionofElectricalandElectronicsEngineersinIsrael,Nov.I.IXIXVI,ppIIIXIX-III0II.
[III]IEEEStandardtestprocedureforthennalevaluationofliquidimmerseddistributionandpowertransfonner,IEEEStdCVVII.I.00-I.IXIXIX,Oct.VII,I.IXIXIX.
[IV]Yann-ChangHuang,Chao-MingHuang,"EvolvingWaveletNetworksforPowerTransfonnerConditionMonitoring,"IEEETrans.PowerDelivery,vol.I.VII,no.II,Apr.II00II,ppAIII-IVI.VI.
[V]Z.Y.Wang,Y.L.Liu,P.I..Griffin,"ACombinedANNandExpertSystemToolforTransfonnerFaultDiagnosis,"IEEETrans.PowerDelivery,vol.I.III,no.IV,Oct.I.IXIXVIII,pp.lIIIIIV--I.IIIIIX.
[VI]M.Dong,"FaultDiagnosisModelPowerTransfonnerBasedonStatisticalLearningTheoryandDissolvedGasAnalysis,"IEEEII00IVInternationalSymposiumonElectricalInsulation,pVIIIV-VIIIVIII.
[VII]B.Suechoey,S.Tadsuan,C.ThammaratandM.Leelajindakrairerk,"Ananalysisoftemperatureandpressureonloadingoil-immerseddistributiontransfonner,"IEEEVIIthInternationalPowerEngineeringConference,Vol.II,pp.VIIIIIV-VIIIIVIII,DecemberII00V.
[VIII]P.Ran,M.H.SunandY.M.Zou,"ZigBeeRoutingSelectionStrategyBasedonDataServicesandEnergy-BalancedZigBeeRouting,"IEEEAsia-PacificConferenceonServicesComputing,Guangdong,Asia,pp.IV00-IV0IV,DecemberII00VI.
[IX]M.Zeghdoud,P.Cordier,M.Terre,"ImpactofClearChannelAssessmentModeonthePerfonnanceofZigBeeOperatinginaWiFiEnvironment,"IEEEWorkshoponOperator-AssistedCommunityNetworks,Berlin,pp.I.-VIII,SeptemberII00VI.
[I.0]w.K.Park,I..HanandW.R.Park,"ZigBeebasedDynamicControlSchemeforMultipleLegacyIRControllableDigitalConsumerDevices,"IEEEConsumerElectronicsSociety,Vol.VIII,No.I,pp.I.VIIII-I.VIIVII,FebruaryII00VII.
[I.I.]K.K.Lee,S.H.Kim,Y.S.ChoiandH.S.Park,"AMeshRoutingProtocolusingClusterLabelintheZigBeeNetwork,"IEEEInternationalConferenceonMobileAdhocandSensorSystems,Vancouver,pp.VIII0I.-VIII0VI,OctoberII00VI.
[I.II]C.H.Lee,"ASurveyonZigBeeTechnologyinSpecificationsandTestPrograms,"ComputersandCommunications,Vol.I.I.IX,pp.IIIII-IVI.,MarchII00VII.
[I.III]S.W.Lwan,IntegrationofRFlDandZigBeenetworkintodevelopinganovelpowermeterandoutagerecorder,MasterThesis,I-ShouUniversity,Taiwan,JuneII00VIII.
[I.IV]S.I..HuangandP.K.Huang,"IntegratedImage-DetectionwithWirelessSensorNetworksforUnmannedSubstationMonitoring",ProceedingsoftheIIVIIIthSymposiumonElectricalPowerEngineering,Taiwan,DecemberII00VII,pp.B0IV.III_I.-V.
[I.V]David.Anderson,DennisJ.Sweeney,ThomasA.Williamsandetai,Statistics(II),TsangHaiBookPublishingCo.,Dec.I.IXIXIX.
Fig.I..Overallconfiguration
Fig.II.Hardwareconfiguration
Fig.III.Firmwareflowchart
Fig.IV.Flowchartofdevicessendingdatatocoordinators
Fig.V.Flowchartofthecoordinatortransmittingdatatoadevice
Fig.VI.FlowchartoftheinterfacedesigninZigBeenetwork
Fig.VII.First-orderlinearregressionmodelwithoneunknownquantity
Fig.VIII.Second-ordernonlinearregressionmodelwithoneunknownquantity
Fig.IX.Third-ordernonlinearregressionmodelwithoneunknownquantity
Fig.I.0.Tangentiallineschematicdiagram
TABLEIANOVAANALYSISTABLE
TABLEIIDECISIONCOEFFICIENTCOMPARISONSAMONGVARIOUSREGRESSIONMODELS
TABLEIIIINTERRELATIONTABLEFORTEMPERATUREANDCURRENT
TABLEIVLOADCURRENTDERIVATIONATVIII0DC
TABLEVLOADCURRENTDERIVATIONATIX0DC
版权保护: 本文由 hbsrm.com编辑，转载请保留链接: www.hbsrm.com/lwqt/wxzs/179.html
