風(fēng)力異步電動(dòng)機(jī)

第第頁(yè)風(fēng)力異步電動(dòng)機(jī)畢業(yè)論文,開(kāi)題報(bào)告。

外文翻譯

2MW風(fēng)力雙饋異步電動(dòng)機(jī)的討論設(shè)計(jì)

指導(dǎo)老師：盛光忠

同學(xué)：安梓銘

〔三峽高?？萍紝W(xué)院〕

畢業(yè)論文,開(kāi)題報(bào)告。

這個(gè)變量速度范圍是成正比的評(píng)級(jí)的轉(zhuǎn)子等通過(guò)變頻器調(diào)速范圍30%[4、5、6、7]轉(zhuǎn)子轉(zhuǎn)換器只需要的DFIG總量的30%的能量而使全面掌握完整的發(fā)電機(jī)輸出功率。這可能導(dǎo)致顯著的成本節(jié)約了轉(zhuǎn)子轉(zhuǎn)換器[4]?；瑒?dòng)環(huán)連接,但需要保持轉(zhuǎn)子繞組,性能安全牢靠。電源發(fā)電機(jī)速度特性,如圖1所示為2MWwind汽輪機(jī)。對(duì)于一個(gè)商業(yè)發(fā)電機(jī)速度隨風(fēng)速,然而這種關(guān)系是為某一特定地點(diǎn)。作為風(fēng)速,并因此機(jī)速度快、輸出功率下降了的風(fēng)力發(fā)電機(jī)減削直至關(guān)閉時(shí)提取風(fēng)是比損失的發(fā)電機(jī)和液力變矩器。操作模式已經(jīng)提出,風(fēng)力機(jī)制造商宣稱(chēng)延伸速度范圍以便在較低的速度能量提取的風(fēng)是比損失在系統(tǒng)等系統(tǒng)能保持聯(lián)系。這個(gè)建議標(biāo)準(zhǔn)的雙(DF)連接在正常運(yùn)用調(diào)速范圍所謂DF異步發(fā)電機(jī)(“模式是用來(lái)延長(zhǎng)低速運(yùn)行。原先的工作已經(jīng)顯示了IG模式能夠運(yùn)作的DFIG滑到80%[8]。這一改變?cè)谶\(yùn)行時(shí)實(shí)現(xiàn)定子從電網(wǎng)DF模式,然后短巡回定子使國(guó)際組操作。全部的發(fā)電機(jī)組轉(zhuǎn)子變頻器在流經(jīng)IG模式。免疫球蛋白曲線相同的曲線為30%DF滑動(dòng)。估量國(guó)際組電力提取的風(fēng)在低速下所獲得的曲線,推斷DF模式。參考扭矩由掌握器(DF和IG模式),就可以很簡(jiǎn)單地來(lái)源于這樣的曲線。扭矩-速度數(shù)據(jù)可以存儲(chǔ)在一個(gè)查表所以參考轉(zhuǎn)矩和轉(zhuǎn)速改變自動(dòng)。

這個(gè)技能的現(xiàn)代DF風(fēng)力渦輪機(jī)不同的無(wú)功功率汲取或產(chǎn)生[6、第九條、第十條]讓風(fēng)渦輪參加無(wú)功功率平衡的格子里。無(wú)功功率在電網(wǎng)的連接中描述的工作,由英國(guó),連接條件小節(jié)CC.6.3.2[11]從國(guó)家電網(wǎng)。無(wú)功要求風(fēng)電場(chǎng)的定義是由圖2。

MVAr點(diǎn)——相當(dāng)于功率因數(shù)為0.95領(lǐng)先于額定兆瓦

MVArB點(diǎn)——相當(dāng)于功率因數(shù)為0.95滯后于額定兆瓦

C-MVAr5點(diǎn)的額定兆瓦

D點(diǎn)-MVAr5%額定兆瓦

E-MVAr12點(diǎn)的額定兆瓦

畢業(yè)論文,開(kāi)題報(bào)告。

方式,在DF)的方式顯示指定。配置程序做了具體的分析,形成了轉(zhuǎn)子的電壓在整個(gè)操作范圍內(nèi)DFIG模式,給出了這種能夠主宰成分浮出水面。這是特別重要的先進(jìn)掌握方案設(shè)計(jì)時(shí)充分概論的工作范圍內(nèi),能被確認(rèn)。仿真模型,它已經(jīng)被證明對(duì)

7.5kW試驗(yàn)室鉆機(jī)[12],是應(yīng)用于現(xiàn)實(shí)的2千瓦風(fēng)力使結(jié)論是關(guān)于擬議中的運(yùn)用IG模式在真實(shí)的風(fēng)力渦輪。

2、連接方法

雙饋異步電機(jī)通常連接如圖3。GSI網(wǎng)格側(cè)逆變器(保持)是一個(gè)固定的直流環(huán)節(jié)電壓與給定的功率因數(shù)的網(wǎng)格(在我們的狀況下,團(tuán)結(jié))。轉(zhuǎn)子側(cè)逆變器(勞損)的掌握,從而使最大能量提取的動(dòng)能的風(fēng)而使定子功率因數(shù)掌握范圍內(nèi)統(tǒng)一要求,盡管網(wǎng)格的功率因數(shù)往往是可取的。另一種連接方式為雙饋電機(jī)如圖4,這叫了異步發(fā)電機(jī)(指定)連接。定子是脫離電網(wǎng)和短路。轉(zhuǎn)子回路圖3。從不變。GSI一樣的掌握方式。DF)目的是為了掌握勞損定子磁鏈在汲取最大功率的動(dòng)能,風(fēng)能。

3、掌握器性能

閉環(huán)掌握方式都和IG模式DF爭(zhēng)論的前期預(yù)備工作[12]但只有一個(gè)7.5億千

畢業(yè)論文,開(kāi)題報(bào)告。

瓦試驗(yàn)室試驗(yàn)平臺(tái)。2千瓦動(dòng)力學(xué)系統(tǒng)會(huì)有所不同,本文爭(zhēng)論了。動(dòng)態(tài)掌握器的性能和IG模式為DF中顯示的是這段2MW風(fēng)力渦輪機(jī)。

3.1DFIG模式(T和Q掌握)

參考價(jià)值的扭矩模式掌握器DF(見(jiàn)圖1)和定子無(wú)功使網(wǎng)格代碼要求達(dá)到

[11],圖2。

畢業(yè)論文,開(kāi)題報(bào)告。

子)的速度如圖7。

穩(wěn)態(tài)Te標(biāo)稱(chēng)值處理的速度、320海里為400轉(zhuǎn)速和4081海里,源自公元1420年轉(zhuǎn)圖1。一個(gè)啟動(dòng)順次需要建立在額定λsr機(jī)器,對(duì)于一個(gè)給定的速度,通過(guò)一段斜坡,圖7,機(jī)器可以產(chǎn)生電力。

一旦該掌握器參考λsr已建立了機(jī)械,特*增加通過(guò)掌握的名義價(jià)值斜坡給定的速度,然后一階躍響應(yīng)50海里在400轉(zhuǎn)速與200海里時(shí)轉(zhuǎn)速適用。公元1420年,該掌握器掌握機(jī)器來(lái)跟蹤Te*果真,參看圖7。

矢量掌握回路的確定值的參考轉(zhuǎn)子電流如圖8。最初的成分快速上升到建立λsr,大約三倍公稱(chēng)穩(wěn)態(tài)值對(duì)于一個(gè)給定的負(fù)荷點(diǎn)。當(dāng)前在額定的限制。最初的解碼器能夠顯著降低,假如一個(gè)較慢的反應(yīng)λsr實(shí)現(xiàn)。

這個(gè)硬中斷懇求優(yōu)先級(jí)別組成,是由扭矩環(huán)使渴望權(quán)力產(chǎn)生。最初有稍微的誤差影響高解碼器的交叉耦合正交循環(huán)系統(tǒng)的條款。一旦名義λsr于機(jī)器徑直和正交環(huán)路的解耦。又一特步引起短暫飆升的硬中斷懇求優(yōu)先級(jí)別*雖然被調(diào)諧到這個(gè)改變是慢于參考價(jià)值。

4、轉(zhuǎn)子的電壓元件

雙方的性能和IG模式DF已經(jīng)在上一節(jié)。兩者都是基于內(nèi)部掌握電流環(huán)和外部掌握回路為轉(zhuǎn)矩和定子無(wú)功功率損耗的案例和轉(zhuǎn)矩和定子磁鏈的IG。再加上解耦方程的PI掌握器的影響,降低產(chǎn)量之間的交叉耦合循環(huán)。最末一部分工作的討論做出貢獻(xiàn)的穩(wěn)態(tài)組件的轉(zhuǎn)子電壓,全部在方程式(1和2),2千瓦機(jī)器來(lái)評(píng)估的重要性,在不同的速度方程式解耦。轉(zhuǎn)子電壓、工具、轉(zhuǎn)子電流、國(guó)稅局,居于

畢業(yè)論文,開(kāi)題報(bào)告。

萬(wàn)物的工具和組件由方程式(1和2)進(jìn)行了DF轉(zhuǎn)速范圍內(nèi)(1000年到1950年轉(zhuǎn)矩確定)的正常從圖表1),和定子動(dòng)力因素、pfs、范圍的0.9落后領(lǐng)先到0.9%。只有pfs被視為GSI可能保持團(tuán)結(jié)酚醛風(fēng)輪變頻器連接到網(wǎng)格的獨(dú)立的勞損。

圖9所示的是改變的速度和vrdqs定子無(wú)功功率范圍的調(diào)查。vrds組件的主導(dǎo)的穩(wěn)態(tài)的ωsfσirqs的壓降和λsq后被忽視的是零組件選擇參考幀。這可以比較圖9和數(shù)字。在一個(gè)2千瓦的vrqs機(jī)床主導(dǎo)下的ωsf(Lm/Ls)λsd期限為低的總泄漏,降低電感、σirds交叉耦合效應(yīng)的術(shù)語(yǔ)和λs取向的λsq構(gòu)件框架設(shè)置為零。在vrqs改變?cè)诤愣ǖ乃俣?并因此轉(zhuǎn)矩)是由于從irds交叉耦合的定子無(wú)功功率調(diào)整,Qs,因此pfs這個(gè)工具vrqs統(tǒng)治級(jí)的組件和對(duì)稱(chēng)1500rpm;thesynchronous速度4-pole機(jī)。這是經(jīng)公園等[13]。

在穩(wěn)態(tài)改變徑直,irds、正交、irqs、轉(zhuǎn)子電流部件對(duì)速度和Qs如圖10。irds元件的功率因數(shù)、調(diào)整定子無(wú),通過(guò)掌握Qs和太少

s組件調(diào)整。irds確定的價(jià)值的比例提供發(fā)電機(jī)無(wú)功功率的定子和轉(zhuǎn)子回路。irds增加越來(lái)越積極的比例從轉(zhuǎn)子回路Q(chēng)同時(shí)減削了問(wèn)從出口到Q的靜定。越來(lái)越消極irds增加問(wèn)從,減削了定子電路的轉(zhuǎn)子的一面,直到Q是由轉(zhuǎn)子出口。Qs隨維持抱負(fù)Te,因此irds組件無(wú)會(huì)持續(xù)pfs在更高的速度。大致上是恒定的irqs元件恒速恒轉(zhuǎn)矩的能量,積極為產(chǎn)生的定位框架和徑直和正交軸排成一線國(guó)稅局的大小是為全部的額定內(nèi)部條件圖10。

其余的這部分說(shuō)明白轉(zhuǎn)子的電壓,vrdqs、穩(wěn)態(tài)部件從方程式(1和2)。這個(gè)Rrsirds術(shù)語(yǔ)及術(shù)語(yǔ)vrdsRrsirqsvrqs僅僅是irdqs,如圖10,攀登通過(guò)后,所以不顯示。

jσωsfirdqs的交叉耦合條件vrdqs如圖11所示，jσωsfirdqs有助于vrds和σωsfirds從vrqs。Σωsfirds由組成隨既速度和定子無(wú)功功率為定子無(wú)功成正比,與轉(zhuǎn)矩對(duì)于一個(gè)給定的定子動(dòng)力因素。σωsfirds隨著年齡的增長(zhǎng)而增長(zhǎng)速度的組件負(fù)載力矩增加如圖1。σωsfirqs組件是主導(dǎo)學(xué)期在vrds組成eqn(1),在不同步性的速度。在極性的結(jié)果ωsf定義和大小的扭矩。irdqs

畢業(yè)論文,開(kāi)題報(bào)告。

大小是由頻率上升而上升，ωsf與總漏電感。圖12說(shuō)明vrdqs由j(Lm/Ls)和ωsf和λsdq組成，(Lm/Ls)ωsfλsq有助于vrds，這個(gè)學(xué)期大致上是零因定位框架。(Lm/Ls)ωsfλsd抑制vrqs的組成，(Lm/Ls)ωsfλsd的外形組成完全由ωsf.決斷。

5、爭(zhēng)論

分析vrds和vrqs組成的可行性是由占統(tǒng)治地位的條款。λs定位框架的結(jié)果λsq和vrds前饋術(shù)語(yǔ)被忽視所以穩(wěn)態(tài)vrds組件的結(jié)果是Rrsirdsσωsfirqs。三種迥乎不同的區(qū)域,然后可以識(shí)別sub-synchronous速度，關(guān)于同步速度，和超同步速度。vrds的瞬態(tài)響應(yīng)的對(duì)于一個(gè)步驟irds*主導(dǎo)下的pσirds.p(Lm/Ls)λsd作為一個(gè)不足掛齒的效果了λsd術(shù)語(yǔ)是恒定的,假設(shè)一個(gè)僵硬的網(wǎng)格。irds*的脈沖一步穩(wěn)態(tài)值影響的vrqs在vrds的穩(wěn)態(tài)條款，vrqs穩(wěn)定的狀態(tài)是由主導(dǎo)下的λsd期限，Vrqs的瞬態(tài)響應(yīng)由irqs*來(lái)的是由pσirqs周期正如步驟irqs最初是高的。p(Lm/Ls)λsq有一個(gè)最大的作用時(shí)λsq約等于零，在vrqs的vrds周期和步驟周期全部的閱歷值改變的irqs*。

6、結(jié)論

畢業(yè)論文,開(kāi)題報(bào)告。

外文原文

DesignStudyofDoubly-FedInductionGenerators

fora2MWWindTurbine

ABSTRACT

Adesignstudyfora2MWcommercialwindturbineispresentedtoillustratetwoconnectionmethodsforastandarddoubly-fedinductionmachinewhichcane*tendthelowspeedrangedownto80%slipwithoutanincreaseintheratingofthepowerelectronicconverter.Thisfare*ceedsthenormal30%lowerlimit.Thelowspeedconnectionisknownasinductiongeneratormodeandthemachineisoperatedwithashortcircuitedstatorwindingwithallpowerflowbeingthroughtherotorcircuit.AtwoloopcascadedPIcontrolschemehasbeendesignedandtunedforeachmode.The

purposeofthispaperistopresentsimulationresultswhichillustratethedynamicperformanceofthecontrollerforbothdoubly-fedinductiongeneratorconnectionmethodsfora2MWwindturbine.Asimpleanalysisoftherotorvoltageforthedoubly-fedconnectionmethodisincludedasthisdemonstratesthedominantcomponentsthatneedtobeconsideredwhendesigningsuchadvancedcontrolstrategies.

Keywords:Doubly-fed,Inductiongenerator,Windturbine

1.INTRODUCTION

Thereiscontinuinginterestinwindturbines,especiallythosewitharatedpowerofmanymegawatts.This

popularityislargelydrivenbybothenvironmentalconcernsandalsotheavailabilityoffossilfuels.Legislationtoencouragethereductionofthesocalledcarbonfootprintiscurrentlyinplaceandsointerestinrenewablesis

currentlyhigh.Windturbinesarestillviewedasawellestablishedtechnologythathasdevelopedfromfi*edspeedwindturbinestothenowpopularvariablespeedtechnologybasedondoubly-fedinductiongenerators(DFIGs).ADFIGwindturbineisvariablespeedwiththerotorconverterbeingcontrolledsothattherotorvoltagephaseandmagnitudeisadjustedtomaintaintheoptimumtorqueandthenecessarystatorpowerfactor[1,2,3].DFIGtechnologyiscurrentlywelldevelopedandiscommonlyusedinwindturbines.ThestatorofaDFIGisdirectlyconnectedtothegridwithapowerelectronicrotorconverterutilisedbetweentherotorwindingandthegrid.Thevariablespeedrangeisproportionaltotheratingoftherotorconverterandsobylimitingthespeedrangeto30%

[4,5,6,7]therotorconverterneedonlyberatedfor30%ofthetotalDFIGpowerwhilstenablingfullcontroloverthefullgeneratoroutputpower.Thiscanresultinsignificantcostsavingsfortherotorconverter[4].Theslipringconnectiontotherotorwindinghowevermustbemaintainedforreliableperformance.

Thepower–generatorspeedcharacteristicshowninfigure1isforacommercial2MWwindturbine.Thegeneratorspeedvarieswithwindspeedhoweverthisrelationissetforaspecificlocation.Aswindspeed,andthereforemachinespeed,fallsthepoweroutputofthegeneratorreducesuntilthewindturbineisswitchedoffwhenthepowere*tractedfromthewindislessthanthelossesofthegeneratorandconverter.Anoperatingmodehasbeenproposedbyawindturbinemanufacturerthatisclaimedtoe*tendthespeedrangesothatatlowerspeedthepowere*tractedfromthewindisgreaterthanthelossesinthesystemandsothesystemcanremainconnected.Thisproposedthatthestandarddoubly-fed(DF)connectionisusedoverthenormalDFspeedrangeandthe

畢業(yè)論文,開(kāi)題報(bào)告。

so-calledinductiongenerator(IG)modeisusedtoe*tendthelowspeedoperation.PreviousworkhasillustratedthatIGmodeenablestheDFIGtooperatedownto80%slip[8].Thischangeinoperationisachievedby

disconnectingthestatorfromthegridinDFmodeandthenshortcircuitingthestatortoenableIGoperation.AllofthegeneratorpowerflowsthroughtherotorconverterinIGmode.TheIGcurveisidenticaltotheDFcurvefor30%slip.TheestimatedIGpowere*tractedfromthewindatlowspeedsisobtainedbye*trapolatingthecurvefortheDFmode.

Thereferencetorquerequiredbybothcontrollers(DFandIGmode)caneasilybederivedfromthiscurve.Thetorque–speeddatacanthenbestoredinalook-uptablesothereferencetorqueisautomaticallyvariedwithspeed.ThecapabilityofmodernDFwindturbinestovarythereactivepowerabsorbedorgenerated[6,9,10]allowsawindturbinetoparticipateinthereactivepowerbalanceofthegrid.Thereactivepoweratthegridconnectionconsideredinthisworkisdescribed,fortheUK,bytheConnectionConditionsSectionCC.6.3.2[11]availablefromtheNationalGrid.Thereactivepowerrequirementforawindfarmisdefinedbyfigure2.

PointA-MVArequivalentfor0.95leadingpowerfactoratratedMW

PointB-MVArequivalentfor0.95laggingpowerfactoratratedMW

PointC-MVAr-5%ofratedMW

PointD-MVAr5%ofratedMW

PointE-MVAr-12%ofratedMW

TheobjectiveofthispaperistoinvestigatethecontrollerperformanceofDFandIGmodefora2MW,690V,4-poleDFIGusingmachineparametersprovidedbythemanufacturer.Thisisfurtherresearchbuildingona

previouspaperwhichdemonstratedthesteady-stateperformanceofthetwomodesofoperation,DFandIGmode

[8].In[8]theauthorsdiscussedthesteady-stateefficiencyforbothconnections.Thesteady-stateperformanceworkillustratedthattherewerebenefitstooperatingthemachineinoneconnectionmethodasopposedtotheother.

Thispapere*aminesthecontrollability(i.e.transientperformance)ofthe2MWwindturbine.Resultsofthefulldynamiccontroller(currentregulation,decouplingequationsandvectorcontrol)inbothDFmodeandIGmodeareshown.AdetailedanalysisofthecomponentsthatformtherotorvoltageoverthefulloperatingrangeinDFIGmodeispresentedasthisenablesthedominantcontrolcomponentstobeidentified.Thisisparticularlyimportantwhendesigningadvancedcontrolschemesasanoverviewoverthefulloperatingrangecanbeidentified.

Simulationmodels,whichhavebeenvalidatedagainsta7.5kWlaboratoryrig[12],areappliedtoarealistic2MWwindturbinetoenableconclusionstobemaderegardingtheproposeduseofIGmodeinarealwindturbine

2.CONNECTIONMETHODS

Doubly-fedinductionmachinesarecommonlyconnectedasshowninfigure3.Thegridsideinverter(GSI)iscontrolledtomaintainafi*eddclinkvoltagewithagivenpowerfactoratthegrid(inourcaseunity).Therotorsideinverter(RSI)iscontrolledsothema*imumenergyise*tractedfromthekineticenergyofthewindwhilstenablingthestatorpowerfactortobecontrolledwithinthelimitsofthegridrequirementsthoughunitypowerfactorisoftendesirable.

Analternativeconnectionmethodforadoubly-fedmachineisshowninfigure4,herecalledtheinductiongenerator(IG)connection.Thestatorisdisconnectedfromthegridandisshort-circuited.Therotorcircuitisunchangedfromfigure3.TheGSIiscontrolledasinDFmode.TheobjectiveoftheRSIistocontrolthestatorflu*linkagewhilee*tractingthema*imumpowerfromthekineticwindenergy.

3.CONTROLLERPERFORMANCE

AclosedloopcontrollerforbothDFmodeandIGmodehasbeendiscussedinpriorwork[12]butonlyfora7.5

畢業(yè)論文,開(kāi)題報(bào)告。

kWlaboratorytestrig.Thedynamicsofa2MWsystemaresomewhatdifferentandareinvestigatedinthispaper.TheperformanceofthedynamiccontrollerforbothDFandIGmodeareshowninthissectionfora2MWwindturbine.

3.1.DFIGMode(TandQControl)

ThereferencevaluesforthecontrollerinDFmodearetorque(seefigure1)andstatorreactivepowertoenablethegridcoderequirement[11]tobeachieved,figure2.Twospeedsareinvestigatedinthissectiontoenablethe

performanceofthecontrollertobeshownbothaboveandbelowthe20%ofratedpowerlimitfromthegridcoderequirements.Anominalgeneratedpowerof320kWisachievedat1150rpm(lessthan20%ofratedpower)andanominalpowerof1.25MWisachievedat1550rpm(greaterthan20%oftheratedpower).Thereferenceandactualtorque,Te,andstatorreactivepower,Qs,areshownforbothspeeds

infigure5.

Thevalueofreferencetorque,Te*,forbothspeedsisthespecificnominaltorqueforagivenspeedcalculatedfromfigure1;2672Nmfor1150rpmand7701Nmfor1550rpm.Astepof200Nmisappliedatbothspeedstoillustratethedynamicresponsetoastepchangeintorque.Thevalueofreferencestatorreactivepower,Qs*,at1150rpmisvariedbetweenthelimitsspecifiedbythegridcoderequirements;initially5%ofthegeneratedpowerwithastepatt=3.5sto+5%ofthegeneratedpower.At1550rpmthestatorpowerfactor,pfs*,isinitially0.95leadingwithastepchangeatt=3stounitypfsandafinalstepatt=4stoa0.95laggingpfs.Thevectorcontrolloopsaretunedforatimeconstantof0.1sand0.9sfortheTeandtheQsloopsrespectively.Thevectorcontrolisdesignedtohaveaslowerbandwidththanthecurrentregulation.

畢業(yè)論文,開(kāi)題報(bào)告。

Theactualrotorcurrentdirect,irds,andquadrature,irqs,componentscorrespondingtofigure5areshowninfigure6.TheeffectofthestepchangeinTe*isapparentontheirqs(thesuperscriptsindicatesthatthevariableisreferredtothestator)ase*pected.Theirqs*componentat1550rpmcontainssmalltransientresponsesatt=3sandt=4sthatareduetothestepchangesintheQsvalue.ThestepchangeinQs*,showninfigure5,causesafastchangeinirds*,figure6,asthereisinitiallyanerrorbetweenthereferenceandactualQsasthecontroltakesashortwhiletorespond.Thecurrentregulationistunedtoensurethatthebandwidthpreventsthecontrollerrespondingtosuchtransientswhilestillachievingasuitablespeedofresponse.TheequationbasedtuningusedtodesignthecontrollergivessimilarvaluesofproportionalandintegralgainsforthecurrentregulationdirectandquadratureloopstothoseusedbyHoldsworthetal[10].

3.2.IGMode(TandFlu*Control)

ThereferencevaluesforthecontrollerinIGmodearestatorflu*linkageandtorque.Twoconditionsare

investigatedforthe2MWgeneratorinIGmode,start-upandtorquestepresponses,at400rpm(minimumIGmodespeed[12])and1420rpm(generatedpoweratthisspeedcorrespondstotheupperpowerratingofrotorconverter,600kW).Thereferenceandactualtorque,Te,andstatorflu*linkage,λsr(thesuperscriptrindicatesthatthevariableisreferredtotherotor),forbothspeedsareshowninfigure7.

Thesteady-stateTeisthenominalvalueforthespeedofoperation,320Nmfor400rpmand4081Nmfor1420rpmderivedfromfigure1.Astart-upsequenceisrequiredtoestablishtheratedλsrinthemachine,foragivenspeed,bymeansofaramp,figure7,beforethemachinecangeneratepower.

Oncethecontrollerreferenceλsrhasbeenestablishedinthemachine,theTe*isincreasedbymeansofacontrolledramptothenominalvalueforagivenspeedandthenastepresponseof50Nmstepat400rpmand200Nmat1420rpmisapplied.ThecontrollerregulatesthemachinetotrackTe*ase*pected,seefigure7.

Thevectorcontrolloopsdeterminethereferencerotorcurrentvaluesthatareshowninfigure8.Theirdcomponentinitiallyincreasesrapidlytoestablishtheλsrandisappro*imately3timesthenominalsteady-statevalueforagivenloadpoint.Thecurrentiswithintheratedlimitatalltimes.Theinitialirdcanbesignificantlyreducedifaslowerresponseofλsrisimplemented.

Theirqcomponentisregulatedbythetorquelooptoenablethedesiredpowertobegenerated.Initiallythereisaslighterrorduetothehighirdwhichaffectsthequadratureloopbythecrosscouplingterms.Oncenominalλsrisestablishedinthemachinethedirectandquadratureloopsaredecoupled.AgainaTestepcausesatransientspikeinirq*thoughthecontrolistunedtobeslowerthanthischangeinreferencevalue.

畢業(yè)論文,開(kāi)題報(bào)告。

4.CONTRIBUTIONOFROTORVOLTAGECOMPONENTS

TheperformanceofbothDFandIGmodehasbeenillustratedintheprevioussection.BothcontrollersarebasedonaninnercurrentloopandanoutercontrolloopfortorqueandstatorreactivepowerintheDFcaseandtorqueandstatorflu*linkageintheIGcase.DecouplingequationswerethenaddedtothePIcontrolleroutputstoreducetheeffectofcrosscouplingbetweentheloops.Thefinalpartofthisworkstudiesthecontributionofthesteadystatecomponentsofrotorvoltage,giveninfullineqns(1and2),fora2MWmachinetoassesstheimportanceofdecouplingequationsatvariousspeeds.Therotorvoltage,vrs,rotorcurrent,irs,andthenon-differentialcomponentsofvrsgivenbyeqns(1and2)areinvestigatedforthefullDFspeedrange(1000to1950rpm)withthenominaltorquedeterminedfromfigure1,andastatorpowerfactor,pfs,rangeof0.9laggingto0.9leading.OnlythepfsisconsideredastheGSIisassumedtomaintainunitypfattherotorconverterconnectiontothegridindependentoftheRSI.

Figure9showsthevariationofvrdqsforthespeedandstatorreactivepowerrangeinvestigated.Thevrdscomponentisdominatedinthesteady-statebytheωsfσirqstermasthevoltagedropacrossRrsisnegligibleandtheλsq

componentiszeroduetothechoiceofreferenceframe.Thiscanbeconfirmedbycomparingfigure9withfigures

11.Thevrqsina2MWmachineisdominatedbytheωsf(Lm/Ls)λsdtermasthelowtotalleakageinductance,σ,reducestheeffectoftheirdscrosscouplingtermandtheλsorientationframesetstheλsqcomponenttozero.Thevariationinvrqsatconstantspeed(andthereforetorque)isduetothecrosscouplingfromtheirdswhichisregulatingthestatorreactivepower,Qs,andthereforepfs.Thevrsmagnitudeisdominatedbythevrqscomponentandissymmetrical1500rpm;thesynchronousspeedfora4-polemachine.ThisisconfirmedbyParketal[13].Thesteady-statevariationinthedirect,irds,andquadrature,irqs,rotorcurrentcomponentswithrespecttospeedandQsisshowninfigure10.Theirdscomponentregulatesthestatorpowerfactor,pfs,bycontrollingQsandtheirdscomponentregulatesTe.Thevalueofirdsdeterminestheproportionofthegeneratorreactivepowersuppliedbythestatorandrotorcircuits.AnincreasinglypositiveirdsincreasestheproportionofQfromtherotorcircuitwhiledecreasingtheQfromthestatoruntilQise*portedbythestator.AnincreasinglynegativeirdsincreasestheQfromthestatorcircuit,reducingtheQfromtherotorsideuntilQise*portedbytherotor.QsincreaseswithTetomaintainthedesiredpfsandsotheirdscomponentwillbehigherforconstantpfsathigherspeeds.Theirqs

componentisappro*imatelyconstantatconstantspeedduetotheconstanttorqueandispositiveforgeneratedpowerduetotheorientationframeandthedirectandquadraturea*isalignment.Theirsmagnitudeiswithintheratedvalueforallconditionsconsideredinfigure10.

Theremainderofthissectionillustratestherotorvoltage,vrdqs,steady-statecomponentsfromeqns(1and2).TheRrsirdsterminvrdsandtheRrsirqsterminvrqsaresimplyirdqs,figure10,scaledbyRrsandsoarenotshown.

Thejσωsfirdqscrosscouplingtermsofvrdqsareshowninfigure11.Thejσωsfirqstermcontributestovrdsandσωsfirdsformspartofvrqs.Theσωsfirdscomponentvarieswithbothspeedandstatorreactivepowerasstatorreactivepowerisproportionaltotorqueforagivenstatorpowerfactor.Theσωsfirdscomponentincreaseswithspeedastheloadtorqueincreases,figure1.Theσωsfirqscomponentisthedominantterminthevrdscomponent,eqn(1),atnon-synchronousspeeds;thepolarityisaresultofωsfandthemagnitudeisdefinedbythetorque.Themagnitudeisirdqsscaledbyslipfrequency,ωsf,andthetotalleakageinductance,σ.Figure12showsthej(Lm/Ls)ωsfλsdq

畢業(yè)論文,開(kāi)題報(bào)告。

componentofvrdqs.The(Lm/Ls)ωsfλsqtermcontributestovrds;thetermisappro*imatelyzeroduetotheorientationframe.The(Lm/Ls)ωsfλsdtermdominatesthevrqscomponent.Theshapeofthe(Lm/Ls)ωsfλsdcomponentisclearlyinfluencedbyωsf.

5.DISCUSSION

Thisanalysisenablesthevrdsandvrqscomponentstobecharacterisedbythedominantterms.TheλsorientationframeresultsintheλsqfeedforwardterminvrdsbeingnegligibleandsothesteadystatevrdscomponentisaresultofRrsirdsσωsfirqs.Threedistinctregionscanthenbeidentified,sub-synchronousspeed(lowirqsduetolowloadsovrdsisappro*imatelyRrsirds),aboutsynchronousspeed(ωsfisaround0sovrdsisappro*imatelyRrsirds)andsupersynchronousspeed(irdsandirqsarecomparableduetohigherloadtorqueandhighstatorpowerfactorsovrdsisappro*imatelyRrsirdsσωsfirqs).Thetransientresponseofvrdsforastepinirds*isdominatedbythepσirds.Thep(Lm/Ls)λsdtermhasanegligibleeffectastheλsdtermisconstantassumingastiffgrid.Anirds*stepaffectsboththesteadystatevalueofvrqsandthesteadystatetermsinvrds.

Thesteadystatevrqscomponentisdominatedbytheλsdterm,confirmedbyHopfenspergeretal[9](withthee*ceptionofsynchronousspeedwhenthesteadystatevrqsisdependentontheRrsirqsterm).Thetransientresponseofvrqstoanirqs*stepisdominatedbythepσirqstermasthedifferentialofthestepchangeinirqsisinitiallyhigh.Thep(Lm/Ls)λsqtermhasanegligibleeffectasλsqisappro*imatelyzero.Thevrdstermandthesteady-statetermsinvrqsalle*perienceachangeinvalueduetotheirqs*step.

6.CONCLUSIONS

ThispaperhasinvestigatedthecontrollerresponsefortheDFandIGmodeconnectionsfora2MWDFIGwindturbine.Themachineparametersforthe2MWmachinewereprovided,foracommerciallyavailableWRIMusedinwindturbines,bythemanufacturer.The2MWmachineparametersusedinthisworkarenotsimplyalinearscalingofpriorworkona7.5kWmachineandsothecharacteristicsarenotidenticalbetweenthetwomachines.Twoareasofanalysishavebeeninvestigatedwithrespecttothe2MWDFIG.E*istingsimulationmodelshavebeenusedtoevaluatethecontrollabilityandsteady-stateandtransientbehaviourofa2MWDFIGinDFandIGmode.TheoutcomeshowsthatIGmodeisacontrollablemodeofoperationwhichwille*tendthelowspeedoperationasrotorvoltagedecreases(asspeedreduces)andsothevoltagelimitoftheIGBTswillberespectedaswillthecurrentandpowerlimitsofthemachineandconverter.Thecompositionoftherotorvoltagewas

investigatedinDFmodeforthe2MWDFIG.ThisshowedhowtheimportanceofthedecouplingequationsontheperformanceoftheDFIGvariedwithspeed.

ACKNOWLEDGEMENTS

TheauthorsaregratefultoFKIIndustrialDrives