chap1仿真与虚拟现实

Chapter1:IntroductionSystemModelingandSimulation-ch.1IntroductionOverview•Simulation•Systems•Models•Evaluation•ConclusionSystemModelingandSimulation-ch.1IntroductionOverview•Simulation–What’ssimulation–Whensimulationsareused–Fourreasonstousesimulation•Systems––System––Stateandstatechanges•Models––Modelcategories––Criteria•Stepsinasimulationstudy•ConclusionSystemsimulation•Givenasystem,howdoyouevaluateitsperformance?•Threeclassicmethods:––Experiments:Useaconcreteexampleofasystemandtrytomeasureitsperformance––Analysis:Constructamathematicalabstractionofthesystemandderiveequationsdescribingthesystem’sperformance––Simulation:Buildamodel(arepresentation)ofthesystem,alongwithitsoperations,andusethismodeltonumericallyevaluatesystemperformance–usuallywiththehelpofcomputers•Here:simulations!Systemsimulation(cont.)•Openquestions––Whatisasystem?––Whatisperformance?––Onwhatdoesperformancedepend?––Whatisamodel?––Whatareoperationsofasystem?––Howtobuildamodel?––Howtonumericallyevaluateit?––Howtointerprettheresultsofsuchanevaluation?What’ssimulationAsimulationisTheimitationoftheoperationofareal-worldprocessorsystemovertime.Atechnique,wherebyamodelofasystem,isrunincompressedtime,toperformexperimentationforanalyzingsystemperformance.Whetherdonebyhandoronacomputer,simulationinvolvesthegenerationofanartificialhistoryofasystem,andtheobservationofthatartificialhistorytodrawinferencesconcerningtheoperatingcharacteristicsoftherealsystem.模拟与仿真•模拟：对某种现象或变化过程的模仿•仿真：用一套以计算机为主体的软硬件设备模仿实际系统的综合技术•技术层面----仿真Whensimulationareused•Usesofsimulationhaveevolvedwithhardware,software•Theearlyyears(1950s-1960s)–Veryexpensive,specializedtooltouse–Requiredbigcomputers,specialtraining–MostlyinFORTRAN(orevenAssembler)–Processingcostashighas$1000/hourforasub-286levelmachineWhensimulationsareused(cont.)•Theformativeyears(1970s-early1980s)–Computersgotfaster,cheaper–Valueofsimulationmorewidelyrecognized–Simulationsoftwareimproved,buttheywerestilllanguagestobelearned,typed,batchprocessed–Oftenusedtocleanup“disasters”inauto,aerospaceindustries•Carplant;heavydemandforcertainmodel•Lineunder-performing•Simulated,problemidentified•Butdemandhaddriedup—simulationwastoolateWhensimulationareused(cont.)•Therecentpast(late1980s)–Microcomputerpower–SoftwareexpandedintoGUIs,animation–Wideracceptanceacrossmoreareas•Traditionalmanufacturingapplications•Services•Healthcare•“Businessprocesses”–Stillmostlyinlargefirms–Oftenasimulationispartofthe“specs”Whensimulationareused(cont.)•Thepresent–Proliferatingintosmallerfirms–Becomingastandardtool–Beingusedearlierindesignphase–Real-timecontrol•Thefuture–Exploitinginteroperabilityofoperatingsystems–Specialized“templates”forindustries,firms–Automatedstatisticaldesign,analysisFourreasonstousesimulation•Tounderstand–Modeltounderstandtheimpactofsolarradiationandgreenhousegasesontheearth’ssurfacetemperature•Toimprove–Modeltoreducewaitingtimeofpatientsinahospitalemergencyroom•Topredictperformance–Simulationofapotentiallayoutforanewfactory•Toguide–SimulationofstudentdemandforcomputersintheA-1buildingExamples•ManufacturingApplications•MilitaryApplications•Logistics,Transportation,andDistributionApplications•BusinessProcessSimulation•HumanSystemsEx1ex2ex3仿真分类•根据模型的种类分类•物理仿真：按照真实系统的物理性质构造系统的物理模型，并在物理模型上进行试验的过程。优点：直观、形象；•缺点：模型改变困难，试验限制多，投资大•数学仿真：对实际系统进行抽象，并将其特性用数学关系加以描述而得到系统的数学模型，对数学模型进行试验的过程称为数学仿真。也称为计算机仿真。•优点：方便、灵活、经济•缺点：数学模型不易建立，受限于系统建模技术。•半实物仿真：将数学模型与物理模型甚至实物联合起来进行试验。简单的建模，复杂的作实物模型，然后联合起来做试验。仿真分类•根据仿真计算机类型分类：•模拟计算机仿真（50-60年代）•数字计算机仿真（计算机仿真指数字计算机仿真）数学模型用计算机程序加以实现，•数字模拟混合仿真（60-70年代，数字计算机仿真技术不成熟）数字计算机和模拟计算机用数模和模数转换装置交换信息仿真分类•根据仿真时钟与实际时钟的比例关系分类：•实时仿真、仿真时钟与实际时钟完全一致，仿真速度=运行速度，•条件：系统中存在物理模型或实物时。如训练仿真器•亚实时仿真、慢于实际时钟•条件：对仿真速度要求不苛刻时。如离线仿真。•超实时仿真：快于实际时钟如：交通系统的仿真。仿真分类根据仿真的设备•逻辑仿真：只用计算机软硬件•含实物仿真•人在回路中的仿真•根据系统模型的特性分类：连续系统仿真、离散时间系统仿真SystemAlmostallinvestigationtaskshavethenotionofa“system”thatistobeinvestigated,characterized•Systemdefinition––AmericanHeritageDictionary:1.Agroupofinteracting,interrelated,orinterdependentelementsformingacomplexwhole,2.Agroupofinteractingmechanicalorelectricalcomponents––Characterizedbyitspartsandtheirinteractions––Systemasawholeisalsodefinedbyitspurposeorfunction•SystemscanbepartoftherealworldorvirtualentitySystem(cont.)•Systemnotiondependsonabstractionlevel––Onahigherlevel,systemcanbepartofanothersystem––Onalowerlevel,partsofasystemusuallycanberegardedassystemsaswell(subsystems)•Systemsarerecursivelystructured––Ingeneral,noatomicparts––Inpractice,atomicpartsarechosendependingonintentionsofusage•Choiceof“system”abstractiondependsonobjectiveorpurposeofusageSystemstate•Inmostsystems,differentmodesorconditionsofbeingcanbedistinguished:Thesystemcanbeindifferentstates•Becauseofrecursivenatureofsystem,partsalso(usually)haveastate•Statecanbeexpressedbythevaluesofasetofvariables––Systemstateistheaggregationofstateoftheparts•StatecanbecontinuousordiscreteSystemstatechangesInteresting(dynamic)systemschangetheirstateovertime––Staticsystemsdonotchangetheirstate,passageoftimeplaysnoroleIncontinuoussystems,thestateofthesystemisaco