




版权说明:本文档由用户提供并上传,收益归属内容提供方,若内容存在侵权,请进行举报或认领
文档简介
ChaosTheory:ABriefWhatexactlyischaos?Thename"chaostheory"comesfromthefactthatthesystemsthatthetheorydescribesareapparentlydisordered,butchaostheoryisreallyaboutfindingtheunderlyingorderinapparentlyrandomdata.Whenwaschaosfirstdiscovered?Thefirsttrueexperimenter haoswasameteorologist,namedEdwardLorenz.In1960,hewasworkingontheproblemofweatherprediction.Hehadacomputersetup,withasetoftwelveequationstomodeltheweather.Itdidn'tpredicttheweatheritself.Howeverthiscomputerprogramdidtheoreticallypredictwhattheweathermightbe.Figure1:Lorenz'periment:thedifferencweenthestartingvaluesofthesecurvesisonly.000127.(IanStewart,DoesGodyDice?TheMathematicsofChaos,pg.141)Onedayin1961,hewantedtoseeaparticularsequenceagain.Tosavetime,hestartedinthemiddleofthesequence,insteadofthebeginning.Heenteredthenumberoffhisprintoutandlefttoletitrun.Whenhecamebackanhourlater,thesequencehadevolveddifferently.Insteadofthesamepatternasbefore,itdivergedfromthepattern,endingupwildlydifferentfromtheoriginal.(Seefigure1.)Eventuallyhefiguredoutwhathappened.Thecomputerstoredthenumberstosixdecimalcesinitsmemory.Tosavepaper,heonlyhaditprintoutthreedecimalces.Intheoriginalsequence,thenumberwas.506127,andhehadonlytypedthefirstthreedigits,.506.Byallconventionalideasofthetime,itshouldhaveworked.Heshouldhavegottenasequenceveryclosetotheoriginalsequence.Ascientistconsidershimselfluckyifhecangetmeasurementswithaccuracytothreedecimalces.Surelythe andfifth,impossibletomeasureusingreasonablemethods,can'thaveahugeeffectontheoutcomeoftheexperiment.Lorenzprovedthisideawrong.Thiseffectcametobeknownasthebutterflyeffect.Theamountofdifferenceinthestartingpointsofthetwocurvesissosmallthatitiscomparabletoabutterflyflapitswings.Theflapofasinglebutterfly'swingtodayproducesatinychangeinthestateoftheatmosphere.Overaperiodoftime,whattheatmosphereactuallydoesdivergesfromwhatitwouldhavedone.So,inamonth'stime,atornadothatwouldhavedevastatedtheIndonesiancoastdoesn'thappen.Ormaybe hatwasn'tgoingtohappen,does.(IanStewart,DoesGodyDice?TheMathematicsofChaos,pg.141)Thisphenomenon,commontochaostheory,isalsoknownassensitivedependenceoninitialconditions.Justasmallchangeintheinitialconditionscandrasticallychangethelong-termbehaviorofasystem.Suchasmallamountofdifferenceinameasurementmightbeconsideredexperimentalnoise,backgroundnoise,oraninaccuracyoftheequipment.Suchthingsareimpossibletoavoidineventhemostisolatedlab.Withastartingnumberof2,thefinalresultcanbeentirelydifferentfromthesamesystemwithastartingvalueof2.000001.Itissimplyimpossibletoachievethislevelofaccuracy-justtryandmeasuresomethingtothenearestmillionthofan Fromthisidea,Lorenzstatedthatitisimpossibletopredicttheweatheraccuray.However,thisdiscoveryledLorenzontootheraspectsofwhateventuallycametobeknownaschaosLorenzstartedtolookforasimplersystemthathadsensitivedependenceoninitialconditions.Hisfirstdiscoveryhadtwelveequations,andhewantedamu oresimpleversionthatstillhadthisattribute.Hetooktheequationsforconvection,andstrippedthemdown,makingthemunrealisticallysimple.Thesystemnolongerhadanythingtodowithconvection,butitdidhavesensitivedependenceonitsinitialconditions,andtherewereonlythreeequationsthistime.Later,itwasdiscoveredthathisequationspreciselydescribedawaterwheel.Atthetop,waterdripssteadilyintocontainershangingonthewheel'srim.Eachcontainerdripssteadilyfromasmallhole.Ifthestreamofwaterisslow,thetopcontainersneverfillfastenoughtoovercomefriction,butifthestreamisfaster,theweightstartstoturnthewheel.Therotationmightbecomecontinuous.Orifthestreamissofastthattheheavycontainersswingallthewayaroundthebottomanduptheotherside,thewheelmightthenslow,stop,andreverseitsrotation,turningfirstonewayandthentheother.(JamesGleick,Chaos-MakingaNewScience,pg.29)Figure2:TheLorenzAttractor(JamesChaos-MakingaNewScience,pg.Theequationsforthissystemalsoseemedtogiverisetoentirelyrandombehavior.However,whenhegraphedit,asurprisingthinghappened.Theoutputalwaysstayedonacurve,adoublespiral.Therewereonlytwokindsoforderpreviouslyknown:asteadystate,inwhichthevariablesneverchange,andperiodicbehavior,inwhichthesystemgoesintoaloop,repeatingitselfindefini y.Lorenz'sequationsweredefini yordered-theyalwaysfollowedaspiral.Theyneversettleddowntoasinglepoint,buts etheyneverrepeatedthesamething,theyweren'tperiodiceither.HecalledtheimagehegotwhenhegraphedtheequationstheLorenzattractor.(Seefigure2)In1963,Lorenzpublishedapaperdescribingwhathehaddiscovered.He ludedtheunpredictabilityoftheweather,anddiscussedthetypesofequationsthatcausedthistypeofbehavior.Unfortuna y,theonlyjournalhewasabletopublishinwasameteorologicaljournal,becausehewasameteorologist,notamathematicianoraphysicist.Asaresult,Lorenz'sdiscoveriesweren'tacknowledgeduntilyearslater,whentheywererediscoveredbyothers.Lorenzhaddiscoveredsomethingrevolutionary;nowhehadtowaitforsome discoverhim.Anothersysteminwhichsensitivedependenceoninitialconditionsisevidentistheflipofacoin.Therearetwovariablesinaflipcoin:howsoonithitstheground,andhowfastitisflip.Theoretically,itshouldbepossibletocontrolthesevariablesentirelyandcontrolhowthecoinwillendup.Inpractice,itisimpossibletocontrolexactlyhowfastthecoinflipstocontrolitenoughtoknowthefinalresultsofthecointoss.Asimilarproblemoccursinecology,andthepredictionofbiologicalpopulations.Theequationwouldbesimpleifpopulationjustrisesindefini y,buttheeffectofpredatorsandalimitedfoodsupplymakethisequation orrect.Thesimplestequationthattakesthisintoaccountisthefollowing:nextyear'spopulation=r*thisyear'spopulation*(1-thisyear'sInthisequation,thepopulationisanumberbetween0and1,where1representsumpossiblepopulationand0represent tion.Risthegrowthrate.Thequestionwas,howdoesthisparame ffecttheequation?Theobviousansweristhatahighgrowthratemeansthatthepopulationwillsettledownatahighpopulation,whilealowgrowthratemeansthatthepopulationwillsettledowntoalownumber.Thistrendistrueforsomegrowthrates,butnotforeveryone.bifurcationdiagramforthepopulationequation.(JamesGleick,Chaos-MakingaNewScience,pg.71)Onebiologist,RobertMay,decidedtoseewhatwouldhappentotheequationasthegrowthratevaluechanges.Atlowvaluesofthegrowthrate,thepopulationwouldsettledowntoasinglenumber.Forinstance,ifthegrowthratevalueis2.7,thepopulationwillsettledownto.6292.Asthegrowthrate reased,thefinalpopulationwould reaseaswell.Then,somethingweirdhappened.Assoonasthegrowthratepassed3,thelinebrokeintwo.Insteadofsettlingdowntoasinglepopulation,itwouldjumpbetweentwodifferentpopulations.Itwouldbeonevalueforoneyear,gotoanothervaluethenextyear,thenrepeatthecycleforever.Raisingthegrowthratealittlemorecausedittojumpbetweenfourdifferentvalues.Astheparameterrosefurther,thelinebifurcated(doubled)again.Thebifurcationscamefas ndfasteruntilsuddenly,chaosappeared.Pastacertaingrowthrate,itbecomesimpossibletopredictthebehavioroftheequation.However,uponcloserinspection,itispossibletoseewhitestrips.Lookingcloseratthesestripsrevealslittlewindowsoforder,wheretheequationgoesthroughthebifurcationsagainbeforereturningtochaos.Thisself-similarity,thefactthatthegraphhasanexactcopyofitselfhiddendeepinside,cametobeanimportantaspectofchaos.AnemployeeofIBM,BenoitMandelbrotwasamathematicianstudyingthisself-similarity.Oneoftheareashewasstudyingwascottonpricefluctuations.Nomatterhowthedataoncottonpriceswasyzed,theresultsdidnotfitthenormaldistribution.Mandelbroteventuallyobtainedalloftheavailabledataoncottonprices,datingbackto1900.WhenheyzedthedatawithIBM'scomputers,henoticedanastonishingThenumbersthatproducedaberrationsfromthepointofviewofnormaldistributionproducedsymmetryfromthepointofviewofscaling.Eachparticularpricechangewasrandomandunpredictable.Butthesequenceofchangeswastonscale:curvesfordailypricechangesandmonthlypricechangesmatchedperfectly. redibly,yzedMandelbrot'sway,thedegreeofvariationhadremainedconstantoveratumultuoussixty-yearperiodthatsawtwoWorldWarsandadepression.(JamesGleick,Chaos-MakingaNewScience,pg.86) yzednotonlycottonprices,butmanyotherphenomenaaswell.Atonepoint,hewaswonderingaboutthelengthofacoastline.Amapofacoastlinewillshowmanybays.However,measuringthelengthofacoastlineoffamapwillmissminorbaysthatweretoosmalltoshowonthemap.Likewise,walkingalongthecoastlinemissesmicroscopicbaysinbetweengrainsofsand.Nomatterhowmuchacoastlineismagnified,therewillbemorebaysvisibleifitismagnifiedmore.Chaos-Figure4:TheKochcurveGleick,Chaos-MakingaNewScience,pg.Onemathematician,HelgevonKoch,capturedthisideainamathematicalconstructioncalledtheKochcurve.TocreateaKochcurve,imagineanequilaltriangle.Tothemiddlethirdofeachside,addanotherequilaltriangle.Keeponaddingnewtrianglestothemiddlepartofeachside,andtheresultisaKochcurve.(Seefigure4)AmagnificationoftheKochcurvelook actlythesameastheoriginal.Itisanotherself-similarfigure.TheKochcurvebringsupaninterestingparadox.Eachtimenewtrianglesareaddedtothefigure,thelengthofthelinegetslonger.However,theinnerareaoftheKochcurveremainslessthantheareaofacircledrawnaroundtheoriginaltriangle.Essentially,itisalineofinfiniengthsurroundingafinitearea.Togetaroundthisdifficulty,mathematiciansinventedfractaldimensions.Fractalcomesfromthewordfractional.ThefractaldimensionoftheKochcurveissomewherearound1.26.Afractionaldimensionisimpossibletoconceive,butitdoesmakesense.TheKochcurveisrougherthanasmoothcurveorline,whichhasonedimension.S eitisrougherandmorecrinkly,itisbet ttakingupspace.However,it'snotasgoodatfillingupspaceasasquarewithtwodimensionsis,s eitdoesn'treallyhaveanyarea.SoitmakessensethatthedimensionoftheKochcurveissomewhereinbetweenthetwo.Fractalhascometomeananyimagethatdisystheattributeofself-similarity.Thebifurcationdiagramofthepopulationequationisfractal.TheLorenzAttractorisfractal.TheKochcurveisfractal.Duringthistime,scientistsfounditverydifficulttogetworkpublishedaboutchaos.S theyhadnotyetshowntherelevancetoreal-worldsituations,mostscientistsdidnotthinktheresultsofexperiments haoswereimportant.Asaresult,eventhoughchaosisamathematicalphenomenon,mostoftheresearchintochaoswasdonebypeopleinotherareas,suchasmeteorologyandecology.Thefieldofchaossproutedupasahobbyforscientistsworkingonproblemsthatmaybehadsomethingtodowithit.Later,ascientistbythenameofFeigenbaumwaslookingatthebifurcationdiagramagain.Hewaslookingathowfastthebifurcationscome.Hediscoveredthattheycome rate.Hecalculateditas4.669.Inotherwords,hediscoveredtheexactscaleatwhichitwasself-similar.Makethediagram4.669timessmaller,anditlookslikethenextregionofbifurcations.Hedecidedtolookatotherequationstoseeifitwaspossibletodetermineascalingfactorforthemaswell.Muchtohissurprise,thescalingfactorwa actlythesame.Notonlywasthiscomplicatedequationdisyingregularity,theregularitywa actlythesameasamuchsimplerequation.Hetriedmanyotherfunctions,andtheyallproducedthesamescalingfactor,4.669.Thiswasarevolutionarydiscovery.Hehadfoundthatawholeclassofmathematicalfunctionsbehavedinthesame,predictableway.Thisuniversalitywouldhelpotherscientistseasilyyzechaoticequations.Universalitygavescientiststhefirsttoolstoyzeachaoticsystem.Nowtheycoulduseasimpleequationtopredicttheoutcomeofamorecomplexequation.Manyscientistswereexploringequationsthatcreatedfractalequations.Themostfamousfractalimageisalsooneofthemostsimple.ItisknownastheMandelbrotset(picturesofthemandelbrotset).Theequationissimple:z=z2+c.ToseeifapointispartoftheMandelbrotset,justtakeacomplexnumberz.Squareit,thenaddtheoriginalnumber.Squaretheresult,thenaddtheoriginalnumber.Repeatthatadinfinitum,andifthenumberkeepsongoinguptoinfinity,itisnotpartoftheMandelbrotset.Ifitstaysdownbelowacertainlevel,itispartoftheMandelbrotset.TheMandelbrotsetistheinnermostsectionofthepicture,andeachdifferentshadeofgrayrepresentshowfaroutthatparticularpointis.OneinterestingfeatureoftheMandelbrotsetisthatthecircularhumpsmatchuptothebifurcationgraph.TheMandelbrotfractalhasthesameself-similarityseenintheotherequations.Infact,zoomingindeepenoughonaMandelbrotfractalwilleventuallyrevealanexactreplicaoftheMandelbrotset,perfectineverydetail.Fractalstructureshavebeennoticedinmanyreal-worldareas,aswellasinmathematician'sminds.Bloodvesselsbranchingoutfurtherandfurther,thebranchesofatree,theinternalstructureofthelungs,graphsofstockmarketdata,andmanyotherreal-worldsystemsallhavesomethingommon:theyareallself-similar.ScientistsatUCSantaCruzfoundchaosinadripwaterfaucet.Byrecordingadripfaucetandrecordingtheperiodsoftime,theydiscoveredthatertainflowvelocity,thedripnolongeroccurredateventimes.Whentheygraphedthedata,theyfoundthatthedripdidindeedfollowapattern.Thehumanheartalsohasachaoticpattern.Thetimweenbeatsdoesnotremainconstant;itdependsonhowmuchactivitya ng,amongotherthings.Under onditions,theheartbeatcanspeedup.Underdifferentconditions,theheartbeatserratically.Itmightevenbecalledachaoticheartbeat.The ysisofaheartbeatcanhelpmedicalresearchersfindwaystoputanabnormalheartbeatbackintoasteadystate,insteadofuncontrolledchaos.Researchersdiscoveredasimplesetofthreeequationsthatgraphedafern.Thisstartedanewidea-perhaps encodesnotexactlywheretheleavesgrow,butaformulathatcontrolstheirdistribution. ,eventhoughitholdsanamazingamountofdata,couldnotholdallofthedatanecessarytodeterminewhereeverycellofthehumanbodygoes.However,byusingfractalformulastocontrolhowthebloodvesselsbranchoutandthenervefibersgetcreated,hasmorethanenoughinformation.Ithasevenbeenspeculatedthatthebrainitselfmightbeorganizedsomehowaccordingtothelawsofchaos.Chaosevenhasapplicationsoutsideofscience.Compu rthasbecomemorerealisticthroughtheuseofchaosandfractals.Now,withasimpleformula,acomputercancreateabeautiful,andrealistictree.Insteadoffollowingaregularpattern,thebarkofatreecanbecreatedaccordingtoaformulathatalmost,butnotquite,repeatsitself.Musiccanbecreatedusingfractalsaswell.UsingtheLorenzattractor,DianaS.Dabby,agraduatestudentinelectricalengineeringattheMassachusettsInstituteofTechnology,hascreatedvariationsofmusicalthemes.("BachtoChaos:ChaoticVariationsonaClassicalTheme",ScienceNews,Dec.24,1994)ByassociatingthemusicalnotesofapieceofmusiclikeBach'sPreludewiththexcoordinatesoftheLorenzattractor,andrunningacomputerprogram,shehascreatedvariationsofthethemeofthesong.Mostmusicianswhohearthenewsoundsbelievethatthevariationsareverymusicalandcreative.Chaoshasalreadyhadalastingeffectonscience,yetthereismuchstilllefttobediscovered.Manyscientistsbelievethattwentiethcenturysciencewillbeknownforonlythreetheories:relativity,quantummechanics,andchaos.Aspectsofchaosshowupeverywherearoundtheworld,fromthecurrentsoftheoceanandtheflowofbloodthroughfractalbloodvesselstothebranchesoftreesandtheeffectsofturbulence.Chaoshasinescapablybecomepartofmodernscience.Aschaoschangedfromalittle-knowntheorytoafullscienceofitsown,ithasreceivedwidespreadpublicity.Chaostheoryhaschangedthedirectionofscience:intheeyesofthegeneralpublic,physicsisnolongersimplythestudyofsubatomicparticlesinabillion-dollarparticleaccelerator,butthestudyofchaoticsystemsandhowtheywork.ChaosFromWikipedia,theThisarticleisaboutchaostheoryinMathematics.ForotherusesofChaostheory,seeChaosTheory(disambiguation).ForotherusesofChaos,seeChaos(disambiguation).Chaostheoryisafieldofstudyinmathematics,withapplicationsinseveraludingphysics,economics,biology,andphilosophy.Chaostheorystudiesthebehaviorofdynamicalsystemsthatarehighlysensitivetoinitialconditions,aneffectwhichispopularlyreferredtoasthebutterflyeffect.Smalldifferencesininitialconditions(suchasthoseduetoroundingerrorsinnumericalcomputation)yieldwidelydivergingoutcomesforchaoticsystems,renderinglong-termpredictionimpossibleingeneral.[1]Thishappenseventhoughthesesystemsaredeterministic,meaningthattheirfuturebehaviorisfullydeterminedbytheirinitialconditions,withnorandomelementsinvolved.[2]Inotherwords,thedeterministicnatureofthesesystemsdoesnotmakethempredictable.[3[4]Thisbehaviorisknownasdeterministicchaos,orsimplychaos.o2.1SensitivitytoinitialDensityofperiodicStrange2.5MinimumcomplexityofachaoticsysDistinguishingrandomfromchaoticCulturalSeeo2.1SensitivitytoinitialDensityofperiodicStrange2.5MinimumcomplexityofachaoticsysDistinguishingrandomfromchaoticCulturalSee oo8.18.28.3Semitechnicalandpopular9ExternalChaostheoryisappliedinmanyscientificdisciplines:mathematics,programming,microbiology,biology,computerdynamics,psychology,androbotics.Chaoticbehaviorhasbeenobservedinthelaboratoryinavarietyof udingcircuits,lasers,oscillatingchemicalreactions,fluiddynamics,andmechanicalandmao-mechanicaldevices,aswellascomputermodelsofchaoticprocesses.Observationsofchaoticbehaviorinnature ludechangesinweather,[5]thedynamicsofsalitesinthesolarsystem,thetimeevolutionofthemaicfieldofcelestialbodies,populationgrowthinecology,thedynamicsoftheactionpotentialsinneurons,andmolecularvibrations.Thereissomecontroversyovertheexistenceofchaoticdynamicsintetectonicsandineconomics.[12][13][14]AsuccessfulapplicationofchaostheoryisinecologywheredynamicalsystemssuchastheRickermodelhavebeenusedtoshowhowpopulationgrowthunderdensitydependencecanleadtochaoticdynamics[citationneeded].Chaostheoryisalsocurrentlybeingappliedtomedicalstudiesofepilepsy,specificallytothepredictionofseeminglyrandomseizuresbyobservinginitialconditions.[15]Quantumchaostheorystudieshowthecorrespondenc weenquantummechanicsandclassicalmechanicsworksinthecontextofchaoticsystems.[16]Recently,anotherfield,calledrelativisticchaos,[17]hasemergedtodescrbesystemsthatfollowthelawsofgeneralrelativity.ThemotionofNstarsinresponsetotheirself-gravity(thegravitationalN-bodyproblem)isgenericallyInelectricalengineering,chaoticsystemsareused ommunications,randomnumbergenerators,andencryptionInnumericalysis,theNewton-Raphsonmethodofapproximatingtherootsofafunctioncanleadtochaotic tionsifthefunctionhasnorealroots.[19][edit]ChaoticThemapdefinedbyThemapdefinedbyx→4x(1–x)andy→x+yifx+y<1(x+y–1otherwise)dis yssensitivitytoinitialconditions.Heretwoseriesofxandyvaluesdivergemarkedlyovertimefromatinyinitialdifference.ommonusage,"chaos"means"astateofdisorder".[20]However, haostheory,thetermisdefinedmoreprecisely.Althoughthereisnouniversallyacceptedmathematicaldefinitionofchaos,acommonlyuseddefinitionsaysthat,foradynamicalsystemtobeclassifiedaschaotic,itmusthavethefollowingproperties:[21]itmustbesensitivetoinitialitmustbetopologicallymixing;itsperiodicorbitsmustbeTherequirementforsensitivedependenceoninitialconditionsimpliesthatthereisasetofinitialconditionsofpositivemeasurewhichdonotconvergetoacycleofanylength. sitivitytoinitialSensitivitytoinitialconditionsmeansthateachpointinsuchasystemisarbitrarilycloselyapproximatedbyotherpointswithsignificantlydifferentfuturetrajectories.Thus,anarbitrarilysmallperturbationofthecurrenttrajectorymayleadtosignificantlydifferentfuturebehaviour.However,ithasbeenshownthatthelasttwopropertiesinthelistaboveactuallyimplysensitivitytoinitialconditions[22][23]andifattentionisrestrictedtointervals,thesecondpropertyimpliestheothertwo[24]( ternative,andingeneralweaker,definitionofchaosusesonlythefirsttwopropertiesintheabovelist[25]).Itisinterestingthatthemostpracticallysignificantcondition,thatofsensitivitytoinitialconditions,isactuallyredund thedefinition,beingimpliedbytwo(orforintervals,one)purelytopologicalconditions,whicharethereforeofgreaterinteresttomathematicians.Sensitivitytoinitialconditionsispopularlyknownasthe"butterflyeffect",socalledbecauseofthetitleofapapergivenbyEdwardLorenzin1972totheAmericanAssociationfortheAdvancementofScienceinWashington,D.C.entitledPredictability:DoestheFlapofaButterfly’sWingsinBrazilsetoffaTorna nTexas?Theflap representsasmallchangeintheinitialconditionofthesystem,whichcausesachainofeventsleadingtolarge-scalephenomena.Hadthebutterflynotflappeditswings,thetrajectoryofthesystemmighthavebeenvastlydifferent.Aconsequenceofsensitivitytoinitialconditionsisthatifwestartwithonlyafiniteamountofinformationaboutthesystem(asisusuallythecaseinpractice),thenbeyondacertaintimethesystemwillnolongerbepredictable.Thisismostfamiliarinthecaseofweather,whichisgenerallypredictableonlyaboutaweekahead.[26]TheLyapunovexponentcharacterisestheextentofthesensitivitytoinitialconditions. twotrajectoriesinphasespacewithinitialseparationdivergeseparationvector.Thus,thereisawholespectrumofLyapunovexponents—thenumberofthemisequaltothenumberofdimensionsofthephasespace.Itiscommontojustrefertothelargestone,i.e.tothe Lyapunovexponent(MLE),becauseitdeterminestheoverallpredictabilityofthesystem.ApositiveMLEisusuallytakenasanindicationthatthesystemischaotic.Therearealsomeasure-theoreticmathematicalconditions(discussedinergodictheory)suchasmixingorbeingaK-systemwhichrelatetosensitivityofinitialconditionsandchaos.[4][edit]TopologicalTopologicalmixing(ortopologicaltransitivity)meansthatthesystemwillevolveovertimesothatanygivenregionoropensetofitsphasespacewilleventuallyoverlapwithanyothergivenregion.Thismathematicalconceptof"mixing"correspondstothestandardintuition,andthemixingofcoloreddyesorfluidsisanexampleofachaoticsystem.Topologicalmixingisoftenomittedfrompopularaccountsofchaos,whichequatechaoswithsensitivitytoinitialconditions.However,sensitivedependenceoninitialconditionsalonedoesnotgivechaos.Forexample,considerthesimpledynamicalsystemproducedbyrepeatedlydoublinganinitialvalue.Thissystemhassensitivedependenceoninitialconditionseverywhere,s eanypairofnearbypointswilleventuallybecomewidelyseparated.However,thi amplehasnotopologicalmixing,andthereforehasnochaos.Indeed,ithasextremelysimplebehaviour:allpoint cept0tendtoinfinity.[edit]DensityofperiodicDensityofperiodicorbitsmeansthateverypointinthespaceisapproachedarbitrarilycloselybyperiodicorbits.Topologicallymixingsystemsfailingthisconditionmaynotdisysensitivitytoinitialconditions,andhencemaynotbechaotic.Forexample,anirrationalrotationofthecircleistopologicallytransitive,butdoesnothavedenseperiodicorbits,andhencedoesnothavesensitivedependenceoninitialconditions.[27]Theorbits.Forexample,→→(orapproximay0.3454915→0.9045085→isan(unstable)orbitofperiod2,andsimilarorbitistforperiods4,8,16,etc.(indeed,foralltheperiodsspecifiedbySharkovskii'stheorem).[28]exhibitsaregularcycleofperiodthreewillalsodisyregularcyclesofeveryotherlengthaswellascompleychaoticorbits.[edit]StrangeUnlikefixed-pointattractorsandlimitcycles,theattractorswhicharisefromchaoticsystems,knownasattractors,havegreatdetailandcomplexity.Strangeattractorsoccurinbothcontinuousdynamicalsystems(suchastheLorenzsystem)andinsomediscretesystems(suchastheHénonmap).Otherdiscretedynamicalfixedpoints–Juliasetscanbethoughtofasstrangerepellers.BothstrangeattractorsandJuliasetstypicallyhaveafractalstructure,andafractaldimensioncanbecalculatedforthem.[edit]MinimumcomplexityofachaoticDiscretechaoticsystems,suchasthelogisticmap,canexhbitstrangeattractorswhatevertheirHowever,thePoaré-Bendixsontheoremshowsthatastrangeattractorcanonlyarisedimensionallinearsystemsareneverchaotic;foradynamicalsystemtodisychaoticbehaviourithastobeeithernonlinear,orinfinite-dimensional.ThePoaré–Bendixsontheoremstatesthatatwodimensionaldifferentialequationhasveryregularbehavior.TheLorenzattractordiscussedaboveisgeneratedbyasystemofthreedifferentialequationswithatotalofseventermsontherighthandside,fiveofwhicharelineartermsandtwoofwhicharequadratic(andthereforenonlinear).Anotherwell-knownchaoticattractorisgeneratedbytheRosslerequationswithseventermsontherighthandside,onlyoneofwhichis(quadratic)nonlinear.Sprott[30]foundathreedimensionalsystemwithjustfivetermsontherighthandside,andwithjustonequadraticnonlinearity,whichexhibitschaosrtainparametervalues.ZhangandHeidel[31][32]showedthat,atleastfordissipativeandconservativequadraticsystems,threedimensionalquadraticsystemswithonlythreeorfourtermsontherighthandsidecannotexhibitchaoticbehavior.Thereasonis,simplyput,thatsolutionstosuchsystemsareasymptotictoatwodimensionalsurfaceandthereforesolutionsarewellbehaved.Whilethe aré–Bendixsontheoremmeans ontinuousdynamicalsystemonEuclideannecannotbechaotic,two-dimensionalcontinuoussystemswithnon-Euclideangeometrycanexhibitchaoticbehaviour.[citationneeded]Perhapssurprisingly,chaosmayoccuralsoinlinearsystems,providedtheyareinfinite-dimensional.[33]Atheoryoflinearchaosisbeingdevelopedinthefunctional ysis,abranchofmathematicalysis.AnearlyproponentofchaostheorywasHenriPo aré.Inthe1880s,whilestudyingthethree-bodyproblem,hefoundthattherecanbeorbitswhicharenonperiodic,andyetnotforeverreasingnorapproachingafixedpoint.[34][35]In1898JacquesHadamardpublishedaninfluentialstudyofthechaoticmotionofaparticleglidingfrictionlesslyonasurfaceofconstantnegativecurvature.[36]Inthesystemstudied,"Hadamard'sbilliards",Hadamardwasabletoshowthatalltrajectoriesareunstableinthatallparticletrajectoriesdivergeexponentiallyfromoneanother,withapositiveLyapunovexponent.Muchoftheearliertheorywasdevelopedalmostentirelybymathematicians,underthenameofergodictheory.Laterstudies,alsoonthetopicofnonlineardifferentialequations,werecarriedoutbyG.D.Birkhoff,[37]A.N.Kolmogorov,[38][39][40]M.L.CartwrightandJ.E.Littlewood,[41]andStephenS
温馨提示
- 1. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
- 2. 本站的文档不包含任何第三方提供的附件图纸等,如果需要附件,请联系上传者。文件的所有权益归上传用户所有。
- 3. 本站RAR压缩包中若带图纸,网页内容里面会有图纸预览,若没有图纸预览就没有图纸。
- 4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
- 5. 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,对用户上传分享的文档内容本身不做任何修改或编辑,并不能对任何下载内容负责。
- 6. 下载文件中如有侵权或不适当内容,请与我们联系,我们立即纠正。
- 7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。
最新文档
- 2024年安徽省安庆望江县某电力公司招聘笔试参考题库附带答案详解
- 2024四川达州水务集团有限公司招用人员20人笔试参考题库附带答案详解
- 2024中国中煤能源股份有限公司海南分公司社会招聘3人笔试参考题库附带答案详解
- 山东省聊城市莘县高中基本能力 中国省区戏剧划分教学实录
- 人教版初中七年级下册历史与社会 5.3.2东部和西部差异显著 教学设计
- 人教版历史(2016)上册教学设计第18课 东晋南北朝时期江南地区的开发
- 八年级英语下册 Unit 10 I've had this bike for three years第一课时 Section A(1a-2d)教学实录(新版)人教新目标版
- 高中数学 5.1.2 瞬时变化率-导数(1)教学设计 苏教版选择性必修第一册
- 2024秋八年级物理上册 第2章 声音与环境 2.1 我们怎样听见声音教学设计(新版)粤教沪版
- 1 负数(教学设计)-2023-2024学年六年级下册数学 人教版
- 中职高教版(2023)语文职业模块-第五单元:走近大国工匠(二)学习工匠事迹 领略工匠风采【课件】
- 2024年山东省济南市中考地理试题卷(含答案解析)
- 2024年太原城市职业技术学院高职单招数学历年参考题库含答案解析
- DB31∕T 795-2014 综合建筑合理用能指南
- GB/T 44979-2024智慧城市基础设施紧凑型城市智慧交通
- 戏剧课程设计方案
- 2025年保密知识试题库附参考答案(精练)
- 物料提升机安全技术操作规程(4篇)
- 临床微生物学检验技术知到智慧树章节测试课后答案2024年秋济宁医学院
- 分级护理质量考核标准
- 图书室管理领导小组及职责
评论
0/150
提交评论