Ch8-Expert-System-Department-of-Computer-Science8專家系統(tǒng)-計算機(jī)科學(xué)系課件_第1頁
Ch8-Expert-System-Department-of-Computer-Science8專家系統(tǒng)-計算機(jī)科學(xué)系課件_第2頁
Ch8-Expert-System-Department-of-Computer-Science8專家系統(tǒng)-計算機(jī)科學(xué)系課件_第3頁
Ch8-Expert-System-Department-of-Computer-Science8專家系統(tǒng)-計算機(jī)科學(xué)系課件_第4頁
Ch8-Expert-System-Department-of-Computer-Science8專家系統(tǒng)-計算機(jī)科學(xué)系課件_第5頁
已閱讀5頁,還剩113頁未讀, 繼續(xù)免費(fèi)閱讀

下載本文檔

版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請進(jìn)行舉報或認(rèn)領(lǐng)

文檔簡介

Ch8ExpertSystemDr.BernardChenPh.D.UniversityofCentralArkansasSpring2019Ch8ExpertSystem1OutlineExpertSystemintroductionRule-BasedExpertSystemGoalDrivenApproachDataDrivenApproachModel-BasedExpertSystemOutlineExpertSystemintroduc2ExpertSystemIntroductionHumanexpertsareabletoperformatasuccessfullevelbecausetheyknowalotabouttheirareasofexpertiseAnExpertSystemuseknowledgespecifictoaproblemdomaintoprovide“expertquality”performanceinthatapplicationareaAswithskilledhumans,expertsystemstendtobespecialists,focusingonanarrowsetofproblemsExpertSystemIntroductionHum3ExpertSystemIntroductionBecauseoftheirheuristic,knowledgeintensivenature,expertsystemsgenerally:SupportinspectionoftheirreasoningprocessesAlloweasymodificationinaddinganddeletingskillsfromknowledgebaseReasonheuristically,usingknowledgetogetusefulsolutionsExpertSystemIntroductionBeca4ExpertSystemIntroductionExpertsystemsarebuilttosolveawiderangeofproblemsindomainsuchasmedicine,math,engineering,chemistry,geology,computerscience,business,low,defenseandeducationTheseprogramsaddressavarietyofproblems,thefollowinglistisasummaryofgeneralexpertsystemproblemcategories:ExpertSystemIntroductionExpe5ExpertSystemIntroductionInterpretation---forminghigh-levelconclusionsfromcollectionsofrawdataPrediction---projectingprobableconsequencesofgivensituationsDiagnosis---determiningthecauseofmalfunctionsbasedonobservablesymptomsExpertSystemIntroductionInte6ExpertSystemIntroductionDesign---findingaconfigurationofsystemcomponentsthatmeetsperformancegoalswhilesatisfyingasetofdesignconstrainsPlanning---devisingasequenceofactionsthatwillachieveasetofgoalsgivenstartingconditionsandruntimeconstrainsExpertSystemIntroductionDesi7TheDesignofRule-BasedExpertSystemarchitectureofatypicalexpertsystemforaparticularproblemdomain.TheDesignofRule-BasedExper8TheDesignofRule-BasedExpertSystemThehearoftheexpertsystemistheknowledgebase,whichcontainstheknowledgeofaparticularapplicationdomainInarule-basedexpertsystem,thisknowledgeismostoftenrepresentedintheformofif…then…Inthefigure,theknowledgebasecontainsbothgeneralandcase-specificinformationTheDesignofRule-BasedExper9TheDesignofRule-BasedExpertSystemTheinferenceengineappliestheknowledgetothesolutionofactualproblemsItisimportanttomaintainthisseparationoftheknowledgeandinferenceenginebecause:MakesitpossibletorepresentknowledgeinamorenaturalfashionExpertsystembuildercanfocusoncapturingandorganizingproblem-solvingknowledgethanthedetailsofcodeimplementationAllowchangetobemadeeasilyAllowsthesamecontrolandinterfacesoftwaretobeusedindifferentsystemsTheDesignofRule-BasedExper10SelectingaproblemExpertSysteminvolveaconsiderableinvestmentofmoneyandhumaneffortResearchershavedevelopedguidelinestodeterminewhetheraproblemisappropriateforexpertsystemsolution:TheneedforthesolutionjustifiesthecostandeffortsofbuildinganexpertsystemHumanexpertiseisnotavailableinallsituationwhereitisneededSelectingaproblemExpertSyst11SelectingaproblemTheproblemdomainiswellstructuredanddoesnotrequirecommonsensereasoningTheproblemmaynotbesolvedusingtraditionalcomputingmethodsCooperativeandarticulateexpertsexistTheproblemispropersizeandscopeSelectingaproblemTheproblem12NASAExampleNASAhassupporteditspresenceinspacebydevelopingafleetofintelligentspaceprobesthatautonomouslyexplorethesolarsystemToachievesuccessthroughyearsintheharshconditionsofspacetravel,acraftneedstobeabletoradicallyreconfigureitscontrolregimeinresponsetofailuresandthenplanaroundthesefailuresduringitremainingflightNASAExampleNASAhassupported13NASAExampleFinally,NASAexpectsthatthesetofpotentialfailurescenariosandpossibleresponseswillbemuchtoolargetousesoftwarethatsupportspreflightenumerationofallcontingenciesLivingstoneisanimplementedkernelforamodel-basedreactiveself-configuringautonomoussystemNASAExampleFinally,NASAexpe14NASAExampleAlong-heldvisionofmodel-basedreasoninghasbeentouseasinglecentralizedmodeltosupportavarietyofengineeringtasksThetasksincludekeeping-trackofdevelopingplansConfirminghardwaremodesReconfiguringhardwareDetectinganomaliesDiagnosisFaultrecoveryNASAExampleAlong-heldvision15NASAExampleNASAExample16NASAExampleItconsistofaheliumtankRegulatorsPropellanttanksApairofmainengineLatchvalvesPyrovalvesNASAExampleItconsistof17NASAExampleTheheliumtankpressurizesthetwopropellanttanks,withtheregulatorsactingtoreducethehighheliumpressureWhenpropellantpathtoamainengineareopen,thepressurizedtankforcesfuelandoxidizerintothemainenginetoproducethrustThepyrovalveistoisolatepartsofthemainenginesubsystemuntiltheyareneeded,ortopermanentlyisolatefailedcomponentsThelatchvalvearecontrolledusingvalvedriversandtheaccelerometerNASAExampleTheheliumtankpr18NASAExampleThrustcanbeprovidedbyeitherofthemainenginesandthereareanumberofwaysofopeningpropellantpathstoeithermainengineNASAExampleThrustcanbeprov19NASAExampleSupposethemainenginesubsystemhasbeenconfiguredtoprovidethrustfromtheleftenginebyopeningthelatchvalvesleadingtoitAndsupposethisenginefails(overheating),sothatisfailstoprovidetherequiredthrustToensurethatthedesirethrustisprovided,thespacecraftmustbetransitionedtoanewconfigurationinwhichthrustisnowprovidedbythemainengineontherightsideNASAExampleSupposethemaine20SelectingaproblemTheprimarypeopleinvolvedinbuildinganexpertsystemaretheknowledgeengineer,domainexpert,andenduserThedomainexpertisprimarilyresponsibleforspellingoutskillstoknowledgeengineerItisoftenusefulforknowledgeengineertobeanoviceintheproblemdomainSelectingaproblemTheprimary21Exploratorydevelopmentcycle

Exploratorydevelopmentcycle22ExploratorydevelopmentcycleItisalsounderstoodthattheprototypemaybethrownawayifitbecomestocumbersomeorifthedesignersdecidetochangetheirbasicapproachtotheproblemAnothermajorfeatureofexpertsystemisthattheprogramneedneverbeconsidered“finished”ExploratorydevelopmentcycleI23OutlineExpertSystemintroductionRule-BasedExpertSystemGoalDrivenApproachDataDrivenApproachModel-BasedExpertSystemOutlineExpertSystemintroduc24StrategiesforstatespacesearchIndatadrivensearch,alsocalledforwardchaining,theproblemsolverbeginswiththegivenfactsoftheproblemandsetoflegalmovesforchangingstateThisprocesscontinuesuntil(wehope!!)itgeneratesapaththatsatisfiesthegoalconditionStrategiesforstatespacesea25“tic-tac-toe”

statespacegraph

“tic-tac-toe”statespacegrap26StrategiesforstatespacesearchAnalternativeapproach(GoalDriven)isstartwiththegoalthatwewanttosolveSeewhatrulescangeneratethisgoalanddeterminewhatconditionsmustbetruetousethemTheseconditionsbecomethenewgoalsWorkingbackwardthroughsuccessivesubgoalsuntil(wehopeagain!)itworkbacktoStrategiesforstatespacesea27Rule-BasedExpertSystemRulebasedexpertsystemrepresentproblem-solvingknowledgeasif…then…ItisoneoftheoldesttechniquesforrepresentingdomainknowledgeinanexpertsystemItisalsooneofthemostnaturalandwidelyusedinpracticalandexperimentalexpertsystemRule-BasedExpertSystemRuleb28Rule-BasedExpertSystemInagoal-drivenexpertsystem,thegoalexpressionisinitiallyplacedinworkingmemoryThesystemmatchesruleconclusionswiththegoal,selectingoneruleandplacingitspremisesintheworkingmemoryThiscorrespondstoadecompositionoftheproblems’goalintosimplersubgoalsTheprocesscontinuesinthenextiterationoftheproductionsystem,withthesepremisesbecomingthenewgoalstomatchRule-BasedExpertSystemInag29AunrealExpertSystemExample

Rule1: if theengineisgettinggas,and theenginewillturnover, then theproblemissparkplugs.Rule2: if theenginedoesnotturnover,and thelightsdonotcomeon then theproblemisbatteryorcables.Rule3: if theenginedoesnotturnover,and thelightsdocomeon then theproblemisthestartermotor.Rule4: if thereisgasinthefueltank,and thereisgasinthecarburetor then theengineisgettinggas.AunrealExpertSystemExample30Theproductionsystematthestartofaconsultationinthecar diagnosticexample.

Theproductionsystematthes31Theproductionsystematthestartofaconsultationinthecar diagnosticexample.Threerulesmatchwiththisexpressioninworkingmemory:rule1,2,and3Ifweresolveconflictsinfavorofthelowest-numberedrule,thenrule1willfireThiscauseXtobeboundtothevaluesparkplugsandthepremisesofrule1tobeplacedintheworkingmemoryTheproductionsystematthes32TheproductionsystemafterRule1hasfired.

TheproductionsystemafterRu33TheproductionsystemafterRule1hasfired.Notethattherearetwopremisestorule1,bothofwhichmustbesatisfiedtoprovetheconclusiontrueSonowweneedtofindoutwhetherTheengineisgettinggas,andTheenginewillturnoverWemaythenfirerule4forwhether“Theengineisgettinggas”TheproductionsystemafterRu34ThesystemafterRule4hasfired.Notethestack-basedapproachtogoalreduction.

ThesystemafterRule4hasfi35Theand/orgraphsearchedinthe

cardiagnosisexample,withtheconclusionofRule4matchingthefirstpremiseofRule1.

Theand/orgraphsearchedint36ExplanationandTransparencyinGoal-DrivenReasoningThefollowingdialoguebeginswiththecomputeraskingtheuseraboutthegoalspresentintheworkingmemory:Gasinfueltank?YESGasincarburetor?YESEnginewillturnover?WHYExplanationandTransparencyi37ExplanationandTransparencyinGoal-DrivenReasoningIngeneral,thetwoquestionsansweredbyrule-basedexpertsystemareWHY?andHOW?WHYmeans“whydidyouaskforthatinformation”TheansweristhecurrentrulethattheproductionsystemisattemptingtofireHOWmeans“Howdidyougettheresult”TheansweristhesequenceofrulesthatwereusedtoconcludeagoalExplanationandTransparencyi38ExplanationandTransparencyinGoal-DrivenReasoningThefollowingdialoguebeginswiththecomputeraskingtheuseraboutthegoalspresentintheworkingmemory:Gasinfueltank?YESGasincarburetor?YESEnginewillturnover?WHYIthasbeenestablishedthat:1.Theengineisgettinggas,2.Theenginewillturnover,(weneedtoknow)Sothatwecanmaketheconclusionthat“Thentheproblemisthesparkplugs.”

ExplanationandTransparencyi39ExplanationandTransparencyinGoal-DrivenReasoningGasinfueltank?YesGasincarburetor?YesEnginewillturnover?Why

Ithasbeenestablishedthat:1.Theengineisgettinggas,2.Theenginewillturnover,Thentheproblemisthesparkplugs.

Howtheengineisgettinggas

Thisfollowsfromrule4:ifgasinfueltank,andgasincarburetorthenengineisgettinggas.gasinfueltankwasgivenbytheusergasincarburetorwasgivenbytheuserExplanationandTransparencyi40OutlineExpertSystemintroductionRule-BasedExpertSystemGoalDrivenApproachDataDrivenApproachModel-BasedExpertSystemOutlineExpertSystemintroduc41Data-DrivenReasoningThepreviousexampleexhibitsgoal-drivensearch.Thesearchwasalsodepth-firstsearchBreadth-firstsearchismorecommoninDataDrivenreasoningThealgorithmforthiscategoryissimple:comparethecontentsofworkingmemorywiththeconditionsofeachruleintherulebaseaccordingtotheorderoftherulesData-DrivenReasoningTheprevi42Data-DrivenReasoningIfapieceofinformationthatmakesupthepremiseofaruleisnottheconclusionofsomeotherrule,thenthatfactwillbedeemed“askable”Forexample:theengineisgettinggasisnotaskableinthepremiseofrule1Data-DrivenReasoningIfapiec43AunrealExpertSystemExample

Rule1: if

(notaskable)theengineisgettinggas,and theenginewillturnover, then theproblemissparkplugs.Rule2: if theenginedoesnotturnover,and thelightsdonotcomeon then theproblemisbatteryorcables.Rule3: if theenginedoesnotturnover,and thelightsdocomeon then theproblemisthestartermotor.Rule4: if thereisgasinthefueltank,and thereisgasinthecarburetor then theengineisgettinggas.AunrealExpertSystemExample44Data-DrivenReasoningData-DrivenReasoning45Data-DrivenReasoningThepremise,theengineisgettinggasisNOTaskable,sorule1failsandcontinuetorule2TheenginedoesnotturnoverisaskableSupposetheanswertothisqueryisfalse,so“theenginewillturnover”isplacedinworkingmemoryData-DrivenReasoningThepremi46TheproductionsystemafterevaluatingthefirstpremiseofRule2,whichthenfails.

Theproductionsystemafterev47TheproductionsystemafterevaluatingthefirstpremiseofRule2,whichthenfails.Rule2fails,sincethefirstoftwoANDpremisesisfalse,wemovetorule3Whererule3alsofailsSofinally,wemovetorule4Theproductionsystemafterev48Thedata-drivenproductionsystemafterconsideringRule4,beginningitssecondpassthroughtherules.

Thedata-drivenproductionsys49Thedata-drivenproductionsystemafterconsideringRule4,beginningitssecondpassthroughtherules.Atthispoint,alltheruleshavebeenconsideredWiththenewcontentsofworkingmemory,weconsidertherulesinorderforthesecondroundThedata-drivenproductionsys50OutlineExpertSystemintroductionRule-BasedExpertSystemGoalDrivenApproachDataDrivenApproachModel-BasedExpertSystemOutlineExpertSystemintroduc51Model-BasedExpertSystemHumanexpertiseisanextremelycomplexcombinationof:TheoreticalknowledgeExperiencedbasedproblemsolvingheuristicsExampleofpastproblemsandtheirsolutionsInterpretiveskillsThroughyearsofexperience,humanexpertdevelopverypowerfulrulesfordealingwithcommonlyencounteredsituationsTheserulesareoftenhighly“complied”Model-BasedExpertSystemHuman52Model-BasedExpertSystemInarule-basedexpertsystemexampleforsemiconductorfailureanalysis,adescriptiveapproachmightbaseon:Discolorationofcomponents(burned-out)HistoryoffaultsinsimilardevicesObservationofcomponentbyelectronmicroscopeHowever,approachesthatuserulestolinkobservationsanddiagnosisdonotofferthebenefitsofadeeperanalysisofdevice’sstructureandfunctionModel-BasedExpertSystemIna53Model-BasedExpertSystemAmorerobust,deeplyexplanatoryapproachwouldbeginwithadetailedmodelofthephysicalstructureofthecircuitandequationsdescribingtheexpectedbehaviorofeachcomponentandtheirinteractions.AknowledgebasedreasonerwhoseanalysisisfoundeddirectlyonthespecificationandfunctionalityofaphysicalsystemiscalledaMODEL-BASEDSystemModel-BasedExpertSystemAmor54Model-BasedExpertSystemThemodelbasedsystemtellsitsuserwhattoexpect,andwhenobservationsdifferfromtheseexpectations,itwillleadtoidentificationoffaultsQualitativemodel-basedreasoningincludes:AdescriptionofeachcomponentinthedeviceAdescriptionofthedevices’internalstructureObservationofthedevices’actualperformanceModel-BasedExpertSystemThem55Model-BasedExpertSystemExampleTheexpectedoutputvaluearegivenin()andtheactualoutputsin[]Model-BasedExpertSystemExam56Model-BasedExpertSystemExampleAtF,wehaveaconflictWecheckthedependenciesatthispointanddeterminedADD1,MULT1andMULT2areinvolvedOneofthesethreedevicesmusthaveafault,sowehavethreehypothesestoconsider:EithertheadderbehaviorisbadoroneofitstwoinputswasincorrectModel-BasedExpertSystemExam57Model-BasedExpertSystemExampleAssumingADD1andoneofitsinputXiscorrect(6)AnotherinputYmustbe(4)Continuethisreasoning,YcannotbeMULT2sinceGiscorrectWeareleftwiththehypothesesthatthefaultliesineitherMULT1orADD1Model-BasedExpertSystemExam58Model-BasedExpertSystemExampleFinally,weshouldnotethatintheexample,therewasassumedtobeasinglefaultydevice.TheworldisnotalwaysthisperfectManyotherpossibleproblemsmayoccur:WireisbrokenFaultyconnectiontothemultiplierModel-BasedExpertSystemExam59Ch8ExpertSystemDr.BernardChenPh.D.UniversityofCentralArkansasSpring2019Ch8ExpertSystem60OutlineExpertSystemintroductionRule-BasedExpertSystemGoalDrivenApproachDataDrivenApproachModel-BasedExpertSystemOutlineExpertSystemintroduc61ExpertSystemIntroductionHumanexpertsareabletoperformatasuccessfullevelbecausetheyknowalotabouttheirareasofexpertiseAnExpertSystemuseknowledgespecifictoaproblemdomaintoprovide“expertquality”performanceinthatapplicationareaAswithskilledhumans,expertsystemstendtobespecialists,focusingonanarrowsetofproblemsExpertSystemIntroductionHum62ExpertSystemIntroductionBecauseoftheirheuristic,knowledgeintensivenature,expertsystemsgenerally:SupportinspectionoftheirreasoningprocessesAlloweasymodificationinaddinganddeletingskillsfromknowledgebaseReasonheuristically,usingknowledgetogetusefulsolutionsExpertSystemIntroductionBeca63ExpertSystemIntroductionExpertsystemsarebuilttosolveawiderangeofproblemsindomainsuchasmedicine,math,engineering,chemistry,geology,computerscience,business,low,defenseandeducationTheseprogramsaddressavarietyofproblems,thefollowinglistisasummaryofgeneralexpertsystemproblemcategories:ExpertSystemIntroductionExpe64ExpertSystemIntroductionInterpretation---forminghigh-levelconclusionsfromcollectionsofrawdataPrediction---projectingprobableconsequencesofgivensituationsDiagnosis---determiningthecauseofmalfunctionsbasedonobservablesymptomsExpertSystemIntroductionInte65ExpertSystemIntroductionDesign---findingaconfigurationofsystemcomponentsthatmeetsperformancegoalswhilesatisfyingasetofdesignconstrainsPlanning---devisingasequenceofactionsthatwillachieveasetofgoalsgivenstartingconditionsandruntimeconstrainsExpertSystemIntroductionDesi66TheDesignofRule-BasedExpertSystemarchitectureofatypicalexpertsystemforaparticularproblemdomain.TheDesignofRule-BasedExper67TheDesignofRule-BasedExpertSystemThehearoftheexpertsystemistheknowledgebase,whichcontainstheknowledgeofaparticularapplicationdomainInarule-basedexpertsystem,thisknowledgeismostoftenrepresentedintheformofif…then…Inthefigure,theknowledgebasecontainsbothgeneralandcase-specificinformationTheDesignofRule-BasedExper68TheDesignofRule-BasedExpertSystemTheinferenceengineappliestheknowledgetothesolutionofactualproblemsItisimportanttomaintainthisseparationoftheknowledgeandinferenceenginebecause:MakesitpossibletorepresentknowledgeinamorenaturalfashionExpertsystembuildercanfocusoncapturingandorganizingproblem-solvingknowledgethanthedetailsofcodeimplementationAllowchangetobemadeeasilyAllowsthesamecontrolandinterfacesoftwaretobeusedindifferentsystemsTheDesignofRule-BasedExper69SelectingaproblemExpertSysteminvolveaconsiderableinvestmentofmoneyandhumaneffortResearchershavedevelopedguidelinestodeterminewhetheraproblemisappropriateforexpertsystemsolution:TheneedforthesolutionjustifiesthecostandeffortsofbuildinganexpertsystemHumanexpertiseisnotavailableinallsituationwhereitisneededSelectingaproblemExpertSyst70SelectingaproblemTheproblemdomainiswellstructuredanddoesnotrequirecommonsensereasoningTheproblemmaynotbesolvedusingtraditionalcomputingmethodsCooperativeandarticulateexpertsexistTheproblemispropersizeandscopeSelectingaproblemTheproblem71NASAExampleNASAhassupporteditspresenceinspacebydevelopingafleetofintelligentspaceprobesthatautonomouslyexplorethesolarsystemToachievesuccessthroughyearsintheharshconditionsofspacetravel,acraftneedstobeabletoradicallyreconfigureitscontrolregimeinresponsetofailuresandthenplanaroundthesefailuresduringitremainingflightNASAExampleNASAhassupported72NASAExampleFinally,NASAexpectsthatthesetofpotentialfailurescenariosandpossibleresponseswillbemuchtoolargetousesoftwarethatsupportspreflightenumerationofallcontingenciesLivingstoneisanimplementedkernelforamodel-basedreactiveself-configuringautonomoussystemNASAExampleFinally,NASAexpe73NASAExampleAlong-heldvisionofmodel-basedreasoninghasbeentouseasinglecentralizedmodeltosupportavarietyofengineeringtasksThetasksincludekeeping-trackofdevelopingplansConfirminghardwaremodesReconfiguringhardwareDetectinganomaliesDiagnosisFaultrecoveryNASAExampleAlong-heldvision74NASAExampleNASAExample75NASAExampleItconsistofaheliumtankRegulatorsPropellanttanksApairofmainengineLatchvalvesPyrovalvesNASAExampleItconsistof76NASAExampleTheheliumtankpressurizesthetwopropellanttanks,withtheregulatorsactingtoreducethehighheliumpressureWhenpropellantpathtoamainengineareopen,thepressurizedtankforcesfuelandoxidizerintothemainenginetoproducethrustThepyrovalveistoisolatepartsofthemainenginesubsystemuntiltheyareneeded,ortopermanentlyisolatefailedcomponentsThelatchvalvearecontrolledusingvalvedriversandtheaccelerometerNASAExampleTheheliumtankpr77NASAExampleThrustcanbeprovidedbyeitherofthemainenginesandthereareanumberofwaysofopeningpropellantpathstoeithermainengineNASAExampleThrustcanbeprov78NASAExampleSupposethemainenginesubsystemhasbeenconfiguredtoprovidethrustfromtheleftenginebyopeningthelatchvalvesleadingtoitAndsupposethisenginefails(overheating),sothatisfailstoprovidetherequiredthrustToensurethatthedesirethrustisprovided,thespacecraftmustbetransitionedtoanewconfigurationinwhichthrustisnowprovidedbythemainengineontherightsideNASAExampleSupposethemaine79SelectingaproblemTheprimarypeopleinvolvedinbuildinganexpertsystemaretheknowledgeengineer,domainexpert,andenduserThedomainexpertisprimarilyresponsibleforspellingoutskillstoknowledgeengineerItisoftenusefulforknowledgeengineertobeanoviceintheproblemdomainSelectingaproblemTheprimary80Exploratorydevelopmentcycle

Exploratorydevelopmentcycle81ExploratorydevelopmentcycleItisalsounderstoodthattheprototypemaybethrownawayifitbecomestocumbersomeorifthedesignersdecidetochangetheirbasicapproachtotheproblemAnothermajorfeatureofexpertsystemisthattheprogramneedneverbeconsidered“finished”ExploratorydevelopmentcycleI82OutlineExpertSystemintroductionRule-BasedExpertSystemGoalDrivenApproachDataDrivenApproachModel-BasedExpertSystemOutlineExpertSystemintroduc83StrategiesforstatespacesearchIndatadrivensearch,alsocalledforwardchaining,theproblemsolverbeginswiththegivenfactsoftheproblemandsetoflegalmovesforchangingstateThisprocesscontinuesuntil(wehope!!)itgeneratesapaththatsatisfiesthegoalconditionStrategiesforstatespacesea84“tic-tac-toe”

statespacegraph

“tic-tac-toe”statespacegrap85StrategiesforstatespacesearchAnalternativeapproach(GoalDriven)isstartwiththegoalthatwewanttosolveSeewhatrulescangeneratethisgoalanddeterminewhatconditionsmustbetruetousethemTheseconditionsbecomethenewgoalsWorkingbackwardthroughsuccessivesubgoalsuntil(wehopeagain!)itworkbacktoStrategiesforstatespacesea86Rule-BasedExpertSystemRulebasedexpertsystemrepresentproblem-solvingknowledgeasif…then…ItisoneoftheoldesttechniquesforrepresentingdomainknowledgeinanexpertsystemItisalsooneofthemostnaturalandwidelyusedinpracticalandexperimentalexpertsystemRule-BasedExpertSystemRuleb87Rule-BasedExpertSystemInagoal-drivenexpertsystem,thegoalexpressionisinitiallyplacedinworkingmemoryThesystemmatchesruleconclusionswiththegoal,selectingoneruleandplacingitspremisesintheworkingmemoryThiscorrespondstoadecompositionoftheproblems’goalintosimplersubgoalsTheprocesscontinuesinthenextiterationoftheproductionsystem,withthesepremisesbecomingthenewgoalstomatchRule-BasedExpertSystemInag88AunrealExpertSystemExample

Rule1: if theengineisgettinggas,and theenginewillturnover, then theproblemissparkplugs.Rule2: if theenginedoesnotturnover,and thelightsdonotcomeon then theproblemisbatteryorcables.Rule3: if theenginedoesnotturnover,and thelightsdocomeon then theproblemisthestartermotor.Rule4: if thereisgasinthefueltank,and thereisgasinthecarburetor then theengineisgettinggas.AunrealExpertSystemExample89Theproductionsystematthestartofaconsultationinthecar diagnosticexample.

Theproductionsystematthes90Theproductionsystematthestartofaconsultationinthecar diagnosticexample.Threerulesmatchwiththisexpressioninworkingmemory:rule1,2,and3Ifweresolveconflictsinfavorofthelowest-numberedrule,thenrule1willfireThiscauseXtobeboundtothevaluesparkplugsandthepremisesofrule1tobeplacedintheworkingmemoryTheproductionsystematthes91TheproductionsystemafterRule1hasfired.

TheproductionsystemafterRu92TheproductionsystemafterRule1hasfired.Notethatthe

溫馨提示

  • 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫網(wǎng)僅提供信息存儲空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時也不承擔(dān)用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。

最新文檔

評論

0/150

提交評論