会员注册 | 登录 | 微信快捷登录 支付宝快捷登录 QQ登录 微博登录 | 帮助中心 人人文库renrendoc.com美如初恋!
站内搜索 百度文库

热门搜索: 直缝焊接机 矿井提升机 循环球式转向器图纸 机器人手爪发展史 管道机器人dwg 动平衡试验台设计

   首页 人人文库网 > 资源分类 > PDF文档下载

外文翻译--机器人在自然地形下爬行 英文版.pdf

  • 资源星级:
  • 资源大小:1.43MB   全文页数:8页
  • 资源格式: PDF        下载权限:注册会员/VIP会员
您还没有登陆,请先登录。登陆后即可下载此文档。
  合作网站登录: 微信快捷登录 支付宝快捷登录   QQ登录   微博登录
友情提示
2:本站资源不支持迅雷下载,请使用浏览器直接下载(不支持QQ浏览器)
3:本站资源下载后的文档和图纸-无水印,预览文档经过压缩,下载后原文更清晰   

外文翻译--机器人在自然地形下爬行 英文版.pdf

ClimbingRobotsinNaturalTerrainTimothyBretl,TeresaMiller,andStephenRockJeanClaudeLatombeAerospaceRoboticsLabRoboticsLaboratoryDepartmentofAeronauticsandAstronauticsComputerScienceDepartmentStanfordUniversity,Stanford,CA94305StanfordUniversity,Stanford,CA94305{tbretl,tgmiller,rock}sunvalley.stanford.edulatombecs.stanford.eduKeywordsMotionplanning,climbing,robotics,leggedrobots,highriskaccess,naturalterrain.AbstractThispaperpresentsageneralframeworkforplanningthequasistaticmotionofclimbingrobots.Theframeworkisinstantiatedtocomputeclimbingmotionsofathreelimbedrobotinverticalnaturalterrain.Anexampleresultingpaththroughalargesimulatedenvironmentispresented.Theplanningproblemisoneoffivefundamentalchallengestothedevelopmentofrealroboticsystemsabletoclimbrealnaturalterrain.Eachofthefourotherareashardwaredesign,control,sensing,andgraspingisalsodiscussed.1IntroductionTheworkdescribedinthispaperispartofanefforttodevelopcriticaltechnologiesthatwillenablethedesignandimplementationofanautonomousrobotabletoclimbverticalnaturalterrain.Toourknowledge,thiscapabilityhasnotbeendemonstratedpreviouslyforroboticsystems.Priorapproacheshavedealtwithartificialterrain,eitherusingspecialgraspse.g.,pegs,magnetsadaptedtotheterrainssurfaceorexploitingspecificpropertiesorfeaturesoftheterraine.g.,ductsandpipes112.Developingthiscapabilitywillfurtherourunderstandingofhowhumansperformsuchcomplextasksasclimbingandscramblinginruggedterrain.Thismayproveusefulinthefuturedevelopmentofsophisticatedroboticsystemsthatwilleitheraidorreplacehumansintheperformanceofaggressivetasksindifficultterrain.Examplesincluderoboticsystemsforsuchmilitaryandcivilianusesassearchandrescue,reconnaissance,andplanetaryexploration.Manyissuesneedtobeaddressedbeforerealrobotscanclimbreal,vertical,naturalterrain.Thispaperconsidersfiveofthemostfundamentaloftheseissueshardwaredesign,control,sensing,planning,andgrasping.Oneoftheseissuesinparticular,themotionplanningproblem,isdescribedinmoredetail.AgeneralframeworkforclimbingrobotsispresentedandthisframeworkisinstantiatedtocomputeclimbingmotionsofthethreelimbedrobotshowninFigure1.Simulationresultsareshownfortherobotinanexampleverticalenvironment.2MotivationTheresultsofresearchinthisareawillbenefitanumberofapplicationsandhaveimplicationsforseveralrelatedresearchareas.2.1ApplicationsThispaperismotivatedbyaneedforroboticsystemscapableofprovidingremoteaccesstohighrisknaturalenvironments.Therearemanyterrestrialapplicationsforthesesystems,suchassearchandrescue,caveexploration,humanassistanceforrockandmountainclimbing,andtacticalurbanmissions.Eachoftheseapplicationsrequiresclimbing,descending,ortraversingsteepslopesandbrokenterrain,andthusinvolvesconsiderablehumanrisk.Severalspaceapplicationscouldalsobenefitfromtheseaggressiveroboticsystems.Forexample,sitesonMarswithpotentiallyhighsciencevaluehavebeenidentifiedonclifffaces13.Often,itisneitherpracticalnorfeasibleforflyingrobotstoaccesstheseFig1.Athreelimbedclimbingrobotmovingverticallyonnaturalsurfaces.locations.Therefore,toreachthesesites,robotsmustclimb,descend,ortraversesteepslopes.Futuregoalsforexplorationonotherplanetarybodiesmayrequireaccesstoequallyruggedterrain.2.2ImplicationsInadditiontofurtheringthedevelopmentofaclimbingrobotforverticalnaturalterrain,theresultsofresearchinthisareacouldprovidefundamentalinsightintoseveralrelatedresearchareas.Forexample,thisstudycouldleadtothedevelopmentofbetterstrategiesforroboticwalkingordexterousmanipulation.Humanclimbersoftencommentonanincreaseinbalanceandanexpandedrangeofmovementineverydayactivityastheybecomemoreproficientatthesport.Thisenhancedmobilityisoftenreferredtoasdiscoveringnewdegreesoffreedom,andisrelatedtotheideaofdiscoveringusefulnewmodesofmobilityforextremelycomplicatedhumanoidrobotsordigitalactors.Also,thedevelopmentofplanningalgorithmsforclimbingrobotscouldleadtoabettersetofcriteriaforthedesignofthesetypesofrobots.Thesealgorithmscouldbeappliedtocandidatedesignsinsimulationtodeterminethecapabilitiesoftheresultingrobots,andthustoselectadesign.3FundamentalIssuesTherearefivefundamentalissuesinvolvedinclimbingsteepnaturalterrainhardwaredesign,control,sensing,grasping,andplanning.Asubstantialamountofworkneedstobedoneineachoftheseareasinordertodeveloparealclimbingrobot.ThissectiondescribesthechallengesinvolvedinthefirstfouroftheseareastheplanningproblemwillbediscussedinmoredetailinSection4.3.1HardwareDesignAgoodhardwaredesigncanincreasetheperformanceoftherobot,andoftencanmakeeachoftheotherfundamentalissueseasiertodealwith.However,pastuseofhardwaresolutionsinmaintainingequilibriumgenerallyresultedinafundamentallimitationontheterrainthatcouldbetraversed.Wheeledroboticsystemshavebeenusedtoascendandtraversenaturalslopesofupto50degrees,todescendslopesofupto75degrees,andtoclimboversmallobstaclesinroughterrain.Thesesystemseitherusesomeformofactiveorrockerbogiesuspensionasin12,1416,oruserappellingasin1.Similarresultshavebeenobtainedusingleggedrappellingrobots3,17andasnakelikerobot4.Theterrainthattheseroverscantraverserobustlyisimpressive,butnoneoftheexistingsystemshasbeenshowntobecapableofclimbingnaturalslopesof90degreesorhigher.Wheeledroversandsnakelikerobotshaveaninherentgraspinglimitationthatpreventstheiruseinascendingsustainednearverticalordescendingsustainedpastverticalnaturalslopes.Existingleggedroboticsystemsdonothavethislimitation,butstillhavebypassedtheissueofmaintainingcontactwiththeslopebyusingrappeltethers.Relianceonthesetethersprohibitsinitialcliffascent,andlimitstheslopegradeoncliffdescenttobelow90degrees.Awidevarietyofrobotscapableofclimbingverticalartificialsurfacesisavailable.Mostoftheserobotsexploitsomepropertyofthesurfaceforeasygrasping.Forexample,someoftheserobotsusesuctioncupsorpermanentmagnetstoavoidslipping58.Otherstakeadvantageoffeaturessuchasbalconyhandrails9orpoles10.However,thesurfacepropertiesthatareexploitedbytheserobotsgenerallyarenotavailableinnaturalterrain.Incontrast,thesimplerhardwaredesignsusedby2,11hadnosuchlimitations.Itisexpectedthatsolutionstotheplanningproblemsuchastheonepresentedinthispaperwillallowbasicnaturalverticalterraintobeclimbedbysimilarsystems,inadditiontotheductsandpipesclimbedbyexistingsystems,andwillsuggestdesignmodificationsforbetterperformance.Futurestudiescouldaddresstheuseofothertypesoftoolsforgraspingverticalnaturalsurfaces,suchastoolsfordrillingboltsorplacingothertypesofgearinrock.Theuseofthesetoolswouldallowmorechallengingclimbstobeaccomplished,inthesamewaythataidhelpshumanclimbers18,19.However,thesetoolsbringanincreaseinweightandcomplexity,slowingmovementandlimitingpotentialapplications.3.2ControlTherearethreeprimarycomponentsofthecontrolproblemforaclimbingrobotmaintenanceofequilibrium,endpointslipcontrol,andendpointforcecontrol.Thesethreecomponentsaretightlyrelated.Inordertomaintainbalance,boththelocationofthecenterofmassoftherobotandtheforcesfromcontactswithnaturalfeaturesmustbecontrolled.Controlofslipatthesecontactsisdirectlyrelatedtothedirectionandmagnitudeofthecontactforces.Existingcontroltechniquessuchasthosebasedontheoperationalspaceformulation20couldformabaselineapproachtothedesignofacontrolarchitectureforaclimbingrobot.However,thesetechniquescouldbeextendedinanumberofdifferentwaystoachievebetterperformance.Forexample,futureresearchmightaddressthedesignofanendpointslipcontrollerthatisstablewithrespecttothecurvatureofacontactsurface,ratherthanwithrespecttoapointcontactonly.3.3SensingForcontrolandgrasping,therobotmustbecapableofsensingtheorientationofitsbodywithrespecttothegravityvector,thelocationofitscenterofmass,therelativelocationofcontactsurfacesfromitslimbendpoints,andtheforcesthatitisexertingatcontactswithnaturalfeatures.Forplanning,therobotmustadditionallybeabletolocatenewholdsandgenerateadescriptionoftheirproperties,possiblyrequiringameasurementoflevelsofslipatcontactpoints.Sensorintegration,inordertoacquireandusethisinformationwithalgorithmsforcontrol,grasping,andplanning,isachallengingproblem.Existingengineeringsolutionsareavailablewhichcanleadtothedevelopmentofabaselineapproachineachcase.Forexample,sensorssuchasthosedescribedin21,22canprovidebasicendpointforceandslipmeasurements,aninertialunitandmagneticcompasscanprovidepositioninformation,anonboardvisionsystemcanprovidearoughcharacterizationofholdlocationsandproperties,andencoderscanprovidethelocationofthecenterofmass.However,theimprovementofeachofthesesensorsintermsofperformance,massreduction,orcostreductionpresentsanopenareaforresearch.AlthoughtheperformanceoftheplanningframeworkthatwillbepresentedinSection4wouldbeimprovedwithbettersensorinformation,itdoesnotdependonaperfectmodeloftheenvironmentapriori.Sincetheframeworkleadstofast,onlineimplementation,planscanbeupdatedtoincorporatenewsensorinformationasitbecomesavailable.3.4GraspingTheperformanceofaclimbingrobotisdependentonitsabilitytograspholds,orfeaturesonasteepnaturalsurface.Ithasalreadybeennotedthatspecializedgraspingschemes,relyingonspecificpropertiesofthesurfacesuchasverysmoothtextures,pegs,orhandles,cannotbeusedforgraspingarbitrarynaturalfeatures.Theproblemsinvolvedingraspingnaturalholdswillbeexaminedfurtherinthissection.Traditionallygraspresearchhasbeeninterestedineitherpickingupanobjectorholdingitimmobilealsocalledfixturing.Researchinthissubjectdatesasfarbackas1876itwasshownthataplanarobjectcouldbeimmobilizedusingaminimumoffourfrictionlesspointconstraints23.Goodoverviewsofmorerecentworkcanbefoundin24,25.Inthisfieldanimportantconceptisforceclosure,definedasagraspthatcanresistallobjectmotionsprovidedthattheendeffectorcanapplysufficientlylargeforcesattheunilateralcontacts.25Nearlyallresearchongraspshasfocusedonselecting,characterizing,andoptimizinggraspsthathavethepropertyofforceclosure.However,forthetaskofclimbingagraspneednotachieveforceclosuretobeausefulgrasp.Forexample,arobotmayfindashelflikeholdveryeffectiveforpullingitselfup,eventhoughthisgraspwouldbecompletelyunabletoresistforcesexertedinotherdirections.Forthisreason,thetechniquesforselecting,characterizing,andoptimizinggraspsmustbeexpandedsignificantlytoapplytoclimbingrobots.Characterizationinvolvesexaminingthedirectionandmagnitudesofforcesandtorquesalsocalledwrenchesthatcanbeexertedbythegrasp.Forexample,foronefingergraspsonpointholds,anadequaterepresentationofthisinformationisafrictioncone,whichwillbeusedfortheplanningalgorithmdescribedinSection4.Theideaofcharacterizationalsoencompassesaqualityfactor.Measuresofgraspqualityhavebeenresearchedextensivelyandarewellreviewedin26.Thisworklistseightdexteritymeasuresthatincludeminimizationofjointangledeviationsandmaximizationofthesmallestsingularvalueofthegraspmatrix.Otherrelevantresearchhasbeendoneusingtheconceptofthewrenchspace.Usingthisconcept,qualityisdefinedasthelargestwrenchspaceballthatcanfitwithintheunitgraspwrenchspace27.Thevolumeofthegraspwrenchspace,orofmorespecializedtaskellipsoids,couldbeusedasaqualitymeasure28.Theseideashavebeenexpandedtoincludelimitingmaximumcontactforceandappliedinagraspsimulatortocomputeoptimalgraspswithvarioushandsin3D29,30.However,theconceptofgraspqualityisilldefinedforgraspsthatdonotprovideforceclosure.Dependingonthedirectionthataclimberwishestogo,differentgraspsmaybeofhigherquality.Furthermore,graspqualitygenerallyincludesaconceptofsecurityorstability,andthistooisilldefinedfornonforceabcdFig.2.Fourdifferenthumanclimbinggrasps,theaopengrip,bcrimp,cfingerlock,anddhandjam.closuregrasps.Again,dependingonthedirectionofappliedforces,thesecurityofagraspmaychange.Theconceptofholdqualitymustbedefinedbeforeusefuloptimizationispossible.Also,anefficientwayoftransmittingthisinformationtoacontrollerorplannerisnecessarytoaccomplishtheclimbingtask.Aqualitativeclassificationofdifferenttypesofgraspsalreadyexistsintheliteratureforhumanclimbers19,31.Inthisclassification,graspsarefirstbrokenintotwocategories,thosemeantforpockets,edges,andotherimperfectionsonotherwiseunbrokenverticalrockfaces,andthosemeantforsustainedverticalcracks.SeveralexamplesofdifferentfaceandcrackgraspsareshowninFigure2.Theliteraturegivesaroughideaofthequalityanduseofeachtypeofgraspintermsofcriteriasuchasaperceivedlevelofsecurity,theamountoftorquethatcanbeexertedonahold,andtheamountoffrictionatthepowerpoint.Notonlyisthisexpertintuitionqualitative,butalsoitisclearthathumanclimbersneedtoperformadditionalgraspplanningforspecificcases.AsputbyLong,Thereareasmanydifferentkindsofholdsastherearewaystograbthem31.However,thisintuitioncanbeusedasastartingpointfordeterminingmeaningfulquantitativecriteriaforgraspselectionandoptimization.Acomparisonoftheclimbingliteraturewithpastworkonroboticgraspplanningrevealsseveralotherfundamentaldifferencesbetweenthetwoapplicationsthatmaybecomeimportantinfutureresearch.Forexample,manyclimbingholdsareverysmall,sothefingersusedinaclimbinggraspoftenhavelargediametersrelativetotheobjecttobegrasped.Literatureonroboticgraspingprimarilyconsidersthecasewherethefingershavesmalldiametersrelativetotheobject.Inaddition,someclimbinggrasps,asmentionedaboveandshowninFigure2,arebasedonjammingfingersinacrack.Thistechniqueisverydifferentfromonearobotmightusetopickupanobject,andrequiresahighdegreeofflexibilityandsmalldegreesoffreedominordertounjamthefingers.Clearly,continuedworkonclimbingrobotseventuallywillleadtotheconsiderationofawealthofnewissuesingrasping.4PlanningTheplanningproblemisthefifthfundamentalchallengeforclimbingrobotsinnaturalterrain.Detailsofthemotionplanningframeworkpresentedinthissectionaregivenin32.4.1ChallengesTheplanningproblemforaclimbingrobotconsistsofgeneratingatrajectorythatmovestherobotthroughaverticalenvironmentwhilemaintainingequilibrium.ThisproblemischallengingevenforhumanclimbersClimbingisdescribedbyLongasasingularabcFig.3.Threedifferenthumanclimbingmoves,theabackstep,bstem,andchighstep.

注意事项

本文(外文翻译--机器人在自然地形下爬行 英文版.pdf)为本站会员(英文资料库)主动上传,人人文库网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。 若此文所含内容侵犯了您的版权或隐私,请立即通知人人文库网([email protected]),我们立即给予删除!

温馨提示:如果因为网速或其他原因下载失败请重新下载,重复下载不扣分。

copyright@ 2015-2017 人人文库网网站版权所有
苏ICP备12009002号-5