自动化专业外文翻译Alicia3爬壁机器人的粘着控制.docx
- 文档编号:28314245
- 上传时间:2023-07-10
- 格式:DOCX
- 页数:11
- 大小:55.21KB
自动化专业外文翻译Alicia3爬壁机器人的粘着控制.docx
《自动化专业外文翻译Alicia3爬壁机器人的粘着控制.docx》由会员分享,可在线阅读,更多相关《自动化专业外文翻译Alicia3爬壁机器人的粘着控制.docx(11页珍藏版)》请在冰豆网上搜索。
自动化专业外文翻译Alicia3爬壁机器人的粘着控制
英语原文:
AdhesionControlfortheAlicia3ClimbingRobot
D.LongoandG.Muscat。
DipartimentodiIngegneriaElettricaElettronicaedeiSistemi,Universit'adegli
StudidiCatania,vialeA.Doria6,95125CataniaItalyAbstract.Climbingrobotsareusefuldevicesthatcanbeadoptedinavarietyofapplicationslikemaintenance,building,inspectionandsafetyintheprocessandconstructionindustries.
ThemaintargetoftheAlicia3robotistoinspectnonporousverticalwallwithanyregardforthematerialofthewall.Tomeetthistarget,apneumatic-likeadhesionforthesystemhasbeenselected.AlsothesystemcanmoveoverthesurfacewithasuitablevelocitybymeansoftwoDCmotorsandovercomesomeobstaclethankstoaspecialcupsealing.
Thisadhesiontechnologyrequiresasuitablecontrollertoimprovesystemreliability.Thisisbecausesmallobstaclespassingunderthecupandwallirregularitycanvarythevalueoftheinternalpressureofthecupputtingtherobotinsomeanomalousworkingconditions.Themethodologiesusedforderivinganaccuratesystemmodelandcontrollerwillbeexplainedandsomeresultwillbepresentedinthiswork.
1Introduction
Climbingrobotscanbeusedtoinspectverticalwallstosearchforpotentialdamageorproblemsonexternalorinternalsurfaceofaboveground/undergroundetrochemicalstoragetanks,concretewallsandmetallicstructures[14].Byusingthissystemascarrier,itwillbepossibletoconductanumberofNDIoverthewallbycarryingsuitableinstrumentation[5,6].Themainapplicationoftheproposedsystemistheautomaticinspectionoftheexternalsurfaceofabovegroundpetrochemicalstoragetankswhereitisveryimportanttoperformperiodicinspections(rateofcorrosion,riskofairorwaterpollution)atdifferentrates,asstandardizedbytheAmericanPetroleumInstitute[7].Thesystemcanbealsoadoptedtoinspectconcretedams.
Whilethesekindsofinspectionsareimportanttopreventecologicaldisastersandrisksforthepeopleworkingaroundtheplant,theseareveryexpensivebecausescaffoldingisoftenrequiredandcanbeverydangerous
Fig.1.TypicaloperatingenvironmentandtheAlicia3robotfortechniciansthathavetoperformtheseinspections.Moreover,forsafetyreasons,plantoperationsmustbestoppedandthetankmustbeemptied,cleanedandventilatedwhenhumanoperatorsareconductinginspections.InFig.1(a)and1(b)typicalenvironmentsforclimbingrobotsareshown.Figure1cshowstheAlicia3robotprototypewhileattachedtoaconcretewallduringasystemtest.
2SystemDescription
TheAliciaIIsystem(thebasicmodulefortheAlicia3system)ismainlycomposedbyacup,anaspirator,twoactuatedwheelsthatusetwoDCmotorswithencodersandgearboxesandfourpassivesteelballswithclearaneetoguaranteeplaincontactofthecuptothewall.Thecupcanslideoverawallbymeansofaspecialsealingthatallowsmaintainingasuitablevacuuminsidethecupandatthesametimecreatingtherightamountoffrictionwithrespectsystemweightandarangeofatargetwallkind.
ThestructureoftheAliciaIImodule,showninFig.2,currentlycomprisesthreeconcentricPVCringsheldtogetherbyanaluminumsdisc.Thebiggerringandthealuminumsdischaveadiameterof30cm.Thesealingsystemisallocatedinthefirsttwoexternalrings.Boththetworingsandthesealingare
Fig.2.StructureoftheAliciaIImoduledesignedtobeeasilyreplaceable,astheywearoffwhiletherobotisrunning.Moreoverthesealingallowstherobotpassingoversmallobstacles(about1cmheight)likescrewsorweldingtraces.Thethirdring(thesmallestone)isused
asabaseforacylinderinwhichacentrifugalairaspiratoranditselectricalmotoraremounted.Theaspiratorisusedtodepressurizethecupformedbytheringsandthesealing,sothewholerobotcanadheretothewalllikeastandardsuctioncup.
Themotor/aspiratorsetisveryrobustandiscapableofworkinginharshenvironments.Thetotalweightofthemoduleis4Kg.
TheAlicia3robotismadewiththethreemoduleslinkedtogetherbymeansoftworodsandaspecialrotationaljoint.Byusingtwopneumaticpistonsitispossibletoriseandtolowereachmoduletoovercomeobstacles.Eachmodulecanberaised15cmwithrespecttothewall,soobstaclesthatare10—2cmheight,canbeeasilyovercame.Thesystemisdesignedtobeabletostayattachedusingonlytwocupswhilethethird,anyofthethree,israisedup.Thetotalweightofthesystemisabout20Kg.
3Electro-PneumaticSystemModel
Byusingthiskindofmovementandsealingmethod,itispossible,duetounexpectedsmallobstaclesonthesurface,tohavesomeairleakageinthecup.Thisleakagecancausetheinternalnegativepressuretoriseupandinthissituationtherobotcouldfalldown.Ontheothersideiftheinternalpressureistoolow(high△p),averybignormalforceisappliedtothesystem.Asaconsequenee,thefrictioncanincreaseinsuchawaytonotallowrobotmovements.Thisproblemcanbesolvedbyintroducingacontrollooptoregulatethepressureinsidethechambertoasuitablevaluetosustainthesystem.TheconsideredopenloopsystemandthemosteasilyaccessiblesystemvariableshasbeenschematizedinFig.3;inthisschemethefirstblockincludestheelectricalandthemechanicalsubsystemandthesecondblockincludesthepneumaticsubsystem.TheusedvariablesaretheMotorvoltagereference(theinputsignalthatfixesthemotorpower)andtheVacuumlevel(thenegativepressureinsidethechamber).
Fig.3.Theopenloopsystemconsidered
Fig.4.1/0variableacquisitionscheme
Sinceitisverydifficulttohaveareliableanalyticalmodelofthatsystem,becauseofthebignumberofparametersinvoIved,ithasbeendecidedtoidentifyablackboxdynamicmodelofthesystembyusinginput/outputmeasurements.Thismodelwasdesignedwithtwopurposes:
tocomputeasuitablecontrolstrategyandtoimplementasimulatorfortuningthecontrolparameters.
Anexperimentalsetupwasrealized,asrepresentedinFig.4,byusingtheDS1102DSPboardfromDspaceinordertogenerateandacquiretheinput/outputvariables.SineetheaspiratorisactuatedbyanACmotor,apowerinterfacehasbeenrealizedinordertotranslateinpowertherefereneesignalforthemotorcomingfromaDACchanneloftheDS1102board.TheoutputsystemvariablehasbeenmeasuredwithapiezoresistivepressuresensorwithasuitableelectronicconditioningblockandacquiredwithoneanaloginputoftheDS1102.ThesoftwarerunningontheDSpaceDSPboard,inthisfirstphasesimplygeneratesanexcitingmotorvoltagereferencesignal(pseudorandom,ramporstepsignals)andacquiresthetwoanaloginputswithasamplingtimeof0.1s,storingthedatainitsinternalSRAM.
TypicalInput/OutputmeasurementsarerepresentedinFig.5andFig.6.Inordertoobtainbetterresultsinsystemmodeling,therelationshipbetweenInputandOutputneedstobeconsideredasnon-linear.ANARXmodelhasbeenusedisintheformof
(1),wherefisanonlinearfunction[8,9].
y(k)=f(u(k),u(k-1),...;y(k-1),y(k-2),...)⑴
Toimplementthiskindofnon-linearity,sometrialshavebeendoneusingNeuro-FuzzyandArtificialNeuralNetwork(ANN)methodologies.Oncethatmodelhasbeentrainedtoasuitablemeansquareerror,ithasbeensimulatedgivingitasinputtherealinputmeasurementonly(infinitesteppredictor)[8].So
(1)canbemodifiedinordertoobtain
(2).
勿(k)=f(u(k),u(k-1),...;欲k-1),勿(k-2),...)
(2)
In
(2),4yistheestimatedsystemoutput.Inordertocomparethesimulationresults,anumberofdescriptorhasbeendefinedandused.Amongthesearemeanerror,quadraticmeanerrorandsomecorrelationindexes.AfirstsetofsimulationforbothmethodologieshasbeendonetofindoutthebestI/Oregressiontermschoice.
3.1Neuro-FuzzyIdentification
Usingthiskindofmethodology,thebestmodelstructurewasfoundtobeintheformof(3).
y(t)=f(u(t),y(t-1))(3)
Oncethebestmodelstructurehasbeenfound,sometrialshavebeenperformedmodifyingthenumberofmembershipfunctions.Thebestresults,comparingtheindexesdescribedabove,havebeenobtainedwith3functionsandinFig.7thesimulationresultshasbeenreported.ThestructureoftheNeuro-FuzzymodelistheANFIS-Sugeno[10].
3.2ANNIdentification
Usingthiskindofmethodology,thebestmodelstructurewasfoundtobeintheformof(4).
y(t)=f(u(t),u(t-1),u(t-2),y(t-1),y(t-2))(4)
Asinglelayerperceptronnetworkhasbeenused.ThetrainingalgorithmisthestandardLevenbergMarquardt.
Oncethebestmodelstructurehasbeenfound,sometrialshavebeenperformedmodifyingthenumberofhiddenneurons.Thebestresults,comparingtheindexesdescribedabove,havebeenobtainedwith7hiddenneuronsandinFig.8thesimulationresultshasbeenreported.
Fromacomparisonbetweenthetwomodelsandtheirrelatedindexes,itcanbeseenthattheNeuro-Fuzzymodelhasbestapproximationperformaneeanduselessinputinformation.Inthenextsection,thismodelwillbeusedassystememulatortotuneandtesttherequiredregulator.
4PressureControlAlgorithm
Onceasystemmodelhasbeenobtained,aclosedloopconfigurationlikethatinFig.9,hasbeenconsidered.
Thetargetofthecontrolalgorithmistoregulatetheinternalvacuumleveltoasuitablevalue(fromsometrials,itwasfixedtoabout10kPa)tosustainthewholesystemanditspayload;themaximumsteadystateerrorallowedwasfixedtolessthan200Pa.Moreoverthetimecons
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- 自动化 专业 外文 翻译 Alicia3 机器人 粘着 控制