外文翻译原文-自动控制的历史

编号：毕业设计外文翻译（原文）学院：机电工程学院专业：电气工程及其自动化学生姓名：范与森学号：1200120309指导教师单位：机电工程学院姓名：王斌职称：讲师2016年05月20日53AHistoryofA4.AHistoryofAutomaticControlChristopherBissellAutomaticcontrol,particularlytheapplicationoffeedback,hasbeenfundamentaltothedevel-opmentofautomation.Itsoriginslieinthelevelcontrol,waterclocks,andpneumatics/hydraulicsoftheancientworld.Fromthe17thcenturyon-wards,systemsweredesignedfortemperaturecontrol,themechanicalcontrolofmills,andtheregulationofsteamengines.Duringthe19thcen-turyitbecameincreasinglyclearthatfeedbacksystemswerepronetoinstability.Astabilitycri-terionwasderivedindependentlytowardstheendofthecenturybyRouthinEnglandandHur-witzinSwitzerland.The19thcentury,too,sawthedevelopmentofservomechanisms,firstforshipsteeringandlaterforstabilizationandautopilots.Theinventionofaircraftadded(literally)anewdimensiontotheproblem.Minorskystheoreti-calanalysisofshipcontrolinthe1920sclarifiedthenatureofthree-termcontrol,alsobeingusedforprocessapplicationsbythe1930s.Basedonservoandcommunicationsengineeringdevel-opmentsofthe1930s,anddrivenbytheneedforhigh-performanceguncontrolsystems,thecoherentbodyoftheoryknownasclassicalcon-trolemergedduringandjustafterWWIIintheUS,UKandelsewhere,asdidcyberneticsideas.Meanwhile,analternativeapproachtodynamicmodelinghadbeendevelopedintheUSSRbasedontheapproachesofPoincarandLyapunov.4.1AntiquityandtheEarlyModernPeriod.534.2StabilityAnalysisinthe19thCentury.564.3Ship,AircraftandIndustrialControlBeforeWWII.574.4Electronics,FeedbackandMathematicalAnalysis.594.5WWIIandClassicalControl:Infrastructure604.6WWIIandClassicalControl:Theory.624.7TheEmergenceofModernControlTheory634.8TheDigitalComputer.644.9TheSocio-TechnologicalContextSince1945.654.10ConclusionandEmergingTrends.664.11FurtherReading.67References.67Informationwasgraduallydisseminated,andstate-spaceormoderncontroltechniques,fuelledbyColdWardemandsformissilecontrolsystems,rapidlydevelopedinbothEastandWest.Theimmediatepost-warperiodwasmarkedbygreatclaimsforautomation,butalsogreatfears,whilethedigitalcomputeropenednewpossibilitiesforautomaticcontrol.4.1AntiquityandtheEarlyModernPeriodFeedbackcontrolcanbesaidtohaveoriginatedwiththefloatvalveregulatorsoftheHellenicandArabworlds4.1.TheywereusedbytheGreeksandArabstocontrolsuchdevicesaswaterclocks,oillampsandwinedispensers,aswellasthelevelofwaterintanks.Thepreciseconstructionofsuchsystemsisstillnotentirelyclear,sincethedescriptionsintheoriginalGreekorArabicareoftenvague,andlackillustrations.ThebestknownGreeknamesareKtsebiosandPhilon(thirdcenturyBC)andHeron(firstcenturyAD)whowereactiveintheeasternMediterranean(Alexandria,Byzantium).ThewaterclocktraditionwascontinuedinPartA454PartADevelopmentandImpactsofAutomationtheArabworldasdescribedinbooksbywriterssuchasAl-Jazari(1203)andIbnal-Sa-ati(1206),greatlyinfluencedbytheanonymousArabauthorknownasPseudo-ArchimedesoftheninthtenthcenturyAD,whomakesspecificreferencetotheGreekworkofHeronandPhilon.FloatregulatorsinthetraditionofHeronwerealsoconstructedbythethreebrothersBanuMusainBaghdadintheninthcenturyAD.ThefloatvalvelevelregulatordoesnotappeartohavespreadtomedievalEurope,eventhoughtransla-tionsexistedofsomeoftheclassicaltextsbytheabovewriters.Itseemsrathertohavebeenreinventeddur-ingtheindustrialrevolution,appearinginEngland,for283030303030303131464629322735373843424240423838393939ACB444126343533454445453738Fig.4.1Meadsspeedregulator(af-ter4.1)example,inthe18thcentury.ThefirstindependentEu-ropeanfeedbacksystemwasthetemperatureregulatorofCorneliusDrebbel(15721633).DrebbelspentmostofhisprofessionalcareeratthecourtsofJamesIandCharlesIofEnglandandRudolfIIinPrague.Drebbelhimselfleftnowrittenrecords,butanumberofcontem-porarydescriptionssurviveofhisinvention.Essentiallyanalcohol(orother)thermometerwasusedtooperateavalvecontrollingafurnaceflue,andhencethetemper-atureofanenclosure4.2.Thedeviceincludedscrewstoalterwhatwewouldnowcallthesetpoint.Iflevelandtemperatureregulationweretwoofthemajorprecursorsofmoderncontrolsystems,thenPartA4.1AHistoryofAutomaticControl4.1AntiquityandtheEarlyModernPeriod55anumberofdevicesdesignedforusewithwindmillspointedthewaytowardsmoresophisticateddevices.Duringthe18thcenturythemillfantailwasdevelopedbothtokeepthemillsailsdirectedintothewindandtoautomaticallyvarytheangleofattack,soastoavoidex-cessivespeedsinhighwinds.Anotherimportantdevicewasthelift-tenter.Millstoneshaveatendencytosep-arateasthespeedofrotationincreases,thusimpairingthequalityofflour.Anumberoftechniquesweredevel-opedtosensethespeedandhenceproducearestoringforcetopressthemillstonesclosertogether.Ofthese,0123456789101112feet00.51234mFig.4.2Boulton&Wattsteamenginewithcentrifugalgovernor(after4.1)perhapsthemostimportantwereThomasMeadsde-vices4.3,whichusedacentrifugalpendulumtosensethespeedandinsomeapplicationsalsotopro-videfeedback,hencepointingthewaytothecentrifugalgovernor(Fig.4.1).Thefirststeamengineswerethereciprocatingen-ginesdevelopedfordrivingwaterpumps;JamesWattsrotaryenginesweresoldonlyfromtheearly1780s.Butittookuntiltheendofthedecadeforthecentrifu-galgovernortobeappliedtothemachine,followingavisitbyWattscollaborator,MatthewBoulton,toPartA4.156PartADevelopmentandImpactsofAutomationtheAlbionMillinLondonwherehesawalift-tenterinactionunderthecontrolofacentrifugalpendu-lum(Fig.4.2).BoultonandWattdidnotattempttopatentthedevice(which,asnotedabove,hadessen-tiallyalreadybeenpatentedbyMead)buttheydidtryunsuccessfullytokeepitsecret.Itwasfirstcopiedin1793andspreadthroughoutEnglandoverthenexttenyearsStabilityAnalysisinthe19thCenturyWiththespreadofthecentrifugalgovernorintheearly19thcenturyanumberofmajorproblemsbecameap-parent.First,becauseoftheabsenceofintegralaction,thegovernorcouldnotremoveoffset:intheterminol-ogyofthetimeitcouldnotregulatebutonlymoderate.Second,itsresponsetoachangeinloadwasslow.Andthirdly,(nonlinear)frictionalforcesinthemech-anismcouldleadtohunting(limitcycling).Anumberofattemptsweremadetoovercometheseproblems:forexample,theSiemenschronometricgovernoref-fectivelyintroducedintegralactionthroughdifferentialgearing,aswellasmechanicalamplification.Otherapproachestothedesignofanisochronousgovernor(onewithnooffset)werebasedoningeniousmechan-icalconstructions,butoftenencounteredproblemsofstability.Neverthelessthe19thcenturysawsteadyprogressinthedevelopmentofpracticalgovernorsforsteamen-ginesandhydraulicturbines,includingspring-loadeddesigns(whichcouldbemademuchsmaller,andoperateathigherspeeds)andrelay(indirect-acting)governors4.6.Bytheendofthecenturygovernorsofvarioussizesanddesignswereavailableforeffec-tiveregulationinarangeofapplications,andanumberofgraphicaltechniquesexistedforsteady-statedesign.Fewengineerswereconcernedwiththeanalysisofthedynamicsofafeedbacksystem.InparallelwiththedevelopmentsintheengineeringsectoranumberofeminentBritishscientistsbecameinterestedingovernorsinordertokeepatelescopedi-rectedataparticularstarastheEarthrotated.AformalanalysisofthedynamicsofsuchasystembyGeorgeBidellAiry,AstronomerRoyal,in18404.7clearlydemonstratedthepropensityofsuchafeedbacksys-temtobecomeunstable.In1868JamesClerkMaxwellanalyzedgovernordynamics,promptedbyanelectri-calexperimentinwhichthespeedofrotationofacoilhadtobeheldconstant.HisresultingclassicpaperOngovernors4.8wasreceivedbytheRoyalSocietyon20February.Maxwellderivedathird-orderlinearmodelandthecorrectconditionsforstabilityintermsofthecoefficientsofthecharacteristicequation.Un-abletoderiveasolutionforhigher-ordermodels,heexpressedthehopethatthequestionwouldgaintheattentionofmathematicians.In1875thesubjectfortheCambridgeUniversityAdamsPrizeinmathemat-icswassetasThecriterionofdynamicalstability.OneoftheexaminerswasMaxwellhimself(prizewin-nerin1857)andthe1875prize(awardedin1877)waswonbyEdwardJamesRouth.Routhhadbeenin-terestedindynamicalstabilityforseveralyears,andhadalreadyobtainedasolutionforafifth-ordersys-tem.Inthepublishedpaper4.9wefindderivedtheRouthversionoftherenownedRouthHurwitzstabilitycriterion.Related,independentworkwasbeingcarriedoutincontinentalEuropeataboutthesametime4.5.AsummaryoftheworkofI.A.VyshnegradskiiinSt.PetersburgappearedintheFrenchComptesRendusdelAcademiedesSciencesin1876,withthefullver-sionappearinginRussianandGermanin1877,andinFrenchin1878/79.Vyshnegradskii(generallytranslit-eratedatthetimeasWischnegradski)transformedathird-orderdifferentialequationmodelofasteamen-HMExGDNFLyFig.4.3Vyshnegradskiisstabilitydiagramwithmodernpolepositions(after4.5)PartA4.2AHistoryofAutomaticControl4.3Ship,AircraftandIndustrialControlBeforeWWII57ginewithgovernorintoastandardform3+x2+y+1=0,wherexandybecameknownastheVyshnegradskiipa-rameters.Hethenshowedthatapointinthexyplanedefinedthenatureofthesystemtransientresponse.Fig-ure4.3showsthediagramdrawnbyVyshnegradskii,towhichtypicalpoleconstellationsforvariousregionsintheplanehavebeenadded.In1893AurelBoreslavStodolaattheFederalPoly-technic,Zurich,studiedthedynamicsofahigh-pressurehydraulicturbine,andusedVyshnegradskiismethodtoassessthestabilityofathird-ordermodel.Amorere-alisticmodel,however,wasseventh-order,andStodolaposedthegeneralproblemtoamathematiciancolleagueAdolfHurwitz,whoverysooncameupwithhisversionoftheRouthHurwitzcriterion4.10.Thetwover-sionswereshowntobeidenticalbyEnricoBompianiin19114.11.Atthebeginningofthe20thcenturythefirstgeneraltextbooksontheregulationofprimemoversappearedinanumberofEuropeanlanguages4.12,13.OneofthemostinfluentialwasTollesRegelungderKraftma-schine,whichwentthroughthreeeditionsbetween1905and19224.14.ThelatereditionsincludedtheHurwitzstabilitycriterion.4.3Ship,AircraftandIndustrialControlBeforeWWIIThefirstshipsteeringenginesincorporatingfeedbackappearedinthemiddleofthe19thcentury.In1873JeanJosephLonFarcotpublishedabookonservomotorsinwhichhenotonlydescribedthevariousdesignsde-velopedinthefamilyfirm,butalsogaveanaccountofthegeneralprinciplesofpositioncontrol.Anotherim-portantmaritimeapplicationoffeedbackcontrolwasingunturretoperation,andhydraulicswerealsoexten-sivelydevelopedfortransmissionsystems.Torpedoes,too,usedincreasinglysophisticatedfeedbacksystemsfordepthcontrolincluding,bytheendofthecentury,gyroscopicaction(Fig.4.4).iammgfdMljkrusoyxvwbctpehzqnFig.4.4TorpedoservomotorasfittedtoWhiteheadtorpedoesaround1900(after4.15)Duringthefirstdecadesofthe20thcenturygyro-scopeswereincreasinglyusedforshipstabilizationandautopilots.ElmerSperrypioneeredtheactivestabilizer,thegyrocompass,andthegyroscopeautopilot,filingvariouspatentsovertheperiod19071914.Sperrysautopilotwasasophisticateddevice:aninnerloopcon-trolledanelectricmotorwhichoperatedthesteeringengine,whileanouterloopusedagyrocompasstosensetheheading.Sperryalsodesignedananticipatortoreplicatethewayinwhichanexperiencedhelms-manwouldmeetthehelm(topreventoversteering);theanticipatorwas,infact,atypeofadaptivecontrol4.16.PartA4.358PartADevelopmentandImpactsofAutomationSperryandhissonLawrencealsodesignedaircraftautostabilizersoverthesameperiod,withtheaddedcomplexityofthree-dimensionalcontrol.Bennettde-scribesthesystemusedinanacclaimeddemonstrationinParisin19144.17:ForthissystemtheSperrysusedfourgyroscopesmountedtoformastabilizedreferenceplatform;atrainofelectrical,mechanicalandpneumaticcomponentsdetectedthepositionoftheaircraftrelativetotheplatformandappliedcorrectionsig-nalstotheaircraftcontrolsurfaces.Thestabilizeroperatedforbothpitchandroll.Thesystem1154749453935154131ba3351535543232137292725181739713Fig.4.5TheStabilog,apneumaticcontrollerprovidingproportionalandintegralaction4.18wasnormallyadjustedtogiveanapproximatelydeadbeatresponsetoastepdisturbance.Thein-corporationofderivativeaction.wasbasedonSperrysintuitiveunderstandingofthebehaviourofthesystem,notonanytheoreticalfoundations.Thesystemwasalsoadaptive.adjustingthegaintomatchthespeedoftheaircraft.Significanttechnologicaladvancesinbothshipandaircraftstabilizationtookplaceoverthenexttwodecades,andbythemid1930sanumberofairlineswereusingSperryautopilotsforlong-distanceflights.However,apartfromthestabilityanalysesdiscussedPartA4.3AHistoryofAutomaticControl4.4Electronics,FeedbackandMathematicalAnalysis59inSect.4.2above,whichwerenotwidelyknownatthistime,therewaslittletheoreticalinvestigationofsuchfeedbackcontrolsystems.Oneoftheearliestsig-nificantstudieswascarriedoutbyNicholasMinorsky,publishedin19224.19.MinorskywasborninRussiain1885(hisknowledgeofRussianprovedtobeimpor-tanttotheWestmuchlater).DuringservicewiththeRussianNavyhestudiedtheshipsteeringproblemand,followinghisemigrationtotheUSAin1918,hemadethefirsttheoreticalanalysisofautomaticshipsteering.Thisstudyclearlyidentifiedthewaythatcontrolac-tionshouldbeemployed:althoughMinorskydidnotusethetermsinthemodernsense,herecommendedanappropriatecombinationofproportional,derivativeandintegralaction.Minorskysworkwasnotwidelydisseminated,however.Althoughhegaveagoodthe-oreticalbasisforclosedloopcontrol,hewaswritinginanageofheroicinvention,whenintuitionandpracticalexperienceweremuchmoreimportantforengineeringpracticethantheoreticalanalysis.Importanttechnologicaldevelopmentswerealsobeingmadeinothersectorsduringthefirstfewdecadesofthe20thcentury,althoughagaintherewaslittletheoreticalunderpinning.Theelectricpowerin-dustrybroughtdemandsforvoltageandfrequencyregulation;manyprocessesusingdrivenrollersre-quiredaccuratespeedcontrol;andconsiderableworkwascarriedoutinanumberofcountriesonsys-temsfortheaccuratepointingofgunsfornavalandanti-aircraftgunnery.Intheprocessindustries,measuringinstrumentsandpneumaticcontrollersofincreasingsophisticationweredeveloped.MasonsSta-bilog(Fig.4.5),patentedin1933,includedintegralaswellasproportionalaction,andbytheendofthedecadethree-termcontrollerswereavailablethatalsoincludedpreactorderivativecontrol.Theoreticalprogresswasslow,however,untiltheadvancesmadeinelectronicsandtelecommunicationsinthe1920sand30sweretranslatedintothecontrolfielddur-ingWWII.4.4Electronics,FeedbackandMathematicalAnalysisTherapidspreadoftelegraphyandthentelephonyfromthemid19thcenturyonwardspromptedagreatdealoftheoreticalinvestigationintothebehaviourofelectriccircuits.OliverHeavisidepublishedpapersonhisop-erationalcalculusoveranumberofyearsfrom1888onwards4.20,butalthoughhistechniquesproducedvalidresultsforthetransientresponseofelectricalnetworks,hewasfiercelycriticizedbycontemporarymathematiciansforhislackofrigour,andultimatelyhewasblackballedbytheestablishment.Itwasnotuntiltheseconddecadeofthe20thcenturythatBromwich,CarsonandothersmadethelinkbetweenHeavisidesoperationalcalculusandFouriermethods,andthusprovedthevalidityofHeavisidestechniques4.21.Thefirstthreedecadesofthe20thcenturysawimportantanalysesofcircuitandfilterdesign,partic-ularlyintheUSAandGermany.HarryNyquistandKarlKpfmllerweretwoofthefirsttoconsidertheproblemofthemaximumtransmissionrateoftele-graphsignals,aswellasthenotionofinformationintelecommunications,andbothwentontoanalyzethegeneralstabilityproblemofafeedbackcircuit4.22.In1928Kpfmlleranalyzedthedynamicsofanau-tomaticgaincontrolelectroniccircuitusingfeedback.Heappreciatedthedynamicsofthefeedbacksystem,buthisintegralequationapproachresultedonlyinaapproximationsanddesigndiagrams,ratherthanarig-orousstabilitycriterion.AtaboutthesametimeintheUSA,HaroldBlackwasdesigningfeedbackamplifiersfortranscontinentaltelephony(Fig.4.6).InafamousepiphanyontheHudsonRiverferryinAugust1927herealizedthatnegativefeedbackcouldreducedistor-tionatthecostofreducingoverallgain.BlackpassedontheproblemofthestabilityofsuchafeedbacklooptohisBellLabscolleagueHarryNyquist,whopub-lishedhiscelebratedfrequency-domainencirclementcriterionin19324.23.Nyquistdemonstrated,usingresultsderivedbyCauchy,thatthekeytostabilityiswhetherornottheopenloopfrequencyresponselocusinthecomplexplaneencircles(inNyquistsoriginalconvention)thepoint1+i0.Oneofthegreatadvan-tagesofthisapproachisthatnoanalyticalformoftheopenloopfrequencyresponseisrequired:asetofmeasureddatapointscanbeplottedwithouttheneedforamathematicalmodel.Anotheradvantageisthat,unliketheRouthHurwitzcriterion,anassess-mentofthetransientresponsecanbemadedirectlyfromtheNyquistplotintermsofgainandphasemargins(howclosethelocusapproachesthecriticalpoint).Blacks1934paperreportinghiscontributiontothedevelopmentofthenegativefeedbackamplifierin-cludedwhatwastobecomethestandardclosed-loopanalysisinthefrequencydomain4.24.PartA4.460PartADevelopmentandImpactsofAutomationFeedbackcircuitbetatwoAmplifiercircuite+n+d(E)ebetatwo(E+N+D)betatwo(E+N+D)E+N+DFig.4.6Blacksfeedbackamplifier(after4.24)Thethirdkeycontributortotheanalysisoffeed-backinelectronicsystemsatBellLabswasHendrikBodewhoworkedonequalizersfromthemid1930s,andwhodemonstratedthatattenuationandphaseshiftwererelatedinanyrealizablecircuit4.25.Thedreamoftelephoneengineerstobuildcircuitswithfastcut-offandlowphaseshiftwasindeedonlyadream.ItwasBodewhointroducedthenotionsofgainandphasemargins,andredrewtheNyquistplotinitsnowconven-tionalformwiththecriticalpointat1+i0.Healsointroducedthefamousstraight-lineapproximationstofrequencyresponsecurvesoflinearsystemsplottedonloglogaxes.Bodepresentedhismethodsinaclassictextpublishedimmediatelyafterthewar4.26.Iftheworkofthecommunicationsengineerswasonemajorprecursorofclassicalcontrol,thentheotherwasthedevelopmentofhigh-performanceservosinthe1930s.Theneedforsuchservoswasgeneratedbytheincreasinguseofanaloguesimulators,suchasnetworkanalysersfortheelectricalpowerindustryanddiffer-entialanalysersforawiderangeofproblems.Bytheearly1930ssix-integratordifferentialanalyserswereinoperationatvariouslocationsintheUSAandtheUK.AmajorcenterofinnovationwasMIT,whereVan-nevarBush,NorbertWienerandHaroldHazenhadallcontributedtodesign.In1934HazensummarizedthedevelopmentsofthepreviousyearsinThetheoryofser-vomechanisms4.27.Headoptednormalizedcurves,andparameterssuchastimeconstantanddampingfac-tor,tocharacterizeservo-response,buthedidnotgivenanystabilityanalysis:althoughheappearstohavebeenawareofNyquistsswork,he(l