简介：BecausemodelswitchingsystemisatypicalformofTakagi-Sugeno(T-S)modelwhichisanuniversalapproximatorofcontinuousnonlinearsystems,wedescribethemodelswitchingsystemasmixedlogicaldynamical(MLD)systemanduseitinmodelpredictivecontrol(MPC)inthispaper.Consideringthateachlocalmodelisonlyvalidineachlocalregion,weaddlocalconstraintstolocalmodels.Thestabilityofproposedmulti-modelpredictivecontrol(MMPC)algorithmisanalyzed,andtheperformanceofMMPCisalsodemonstratedonaninulti-multi-output(MIMO)simulatedpHneutralizationprocess.

简介：ThetraditionalGaussianMixtureModel(GMM)forpatternrecognitionisanunsupervisedlearningmethod.Theparametersinthemodelarederivedonlybythetrainingsamplesinoneclasswithouttakingintoaccounttheeffectofsampledistributionsofotherclasses,hence,itsrecognitionaccuracyisnotidealsometimes.ThispaperintroducesanapproachforestimatingtheparametersinGMMinasupervisingway.TheSupervisedLearningGaussianMixtureModel(SLGMM)improvestherecognitionaccuracyoftheGMM.Anexperimentalexamplehasshownitseffectiveness.TheexperimentalresultshaveshownthattherecognitionaccuracyderivedbytheapproachishigherthanthoseobtainedbytheVectorQuantization(VQ)approach,theRadialBasisFunction(RBF)networkmodel,theLearningVectorQuantization(LVQ)approachandtheGMM.Inaddition,thetrainingtimeoftheapproachislessthanthatofMultilayerPerceptrom(MLP).

简介：

简介：Despiteitsname,e-commerceisnotfullyautomated.Typically,onlinetransactionsrequireasignificantlevelofhumanintervention.Customersundertakethesearchforproducts,services,potentialvendorsandbusinesspartners.Theyevaluatealternatives,decidewhatgoodstobuyandwhen,fromwhichvendor,andhowmuchtheyarewillingtopayfor.Theyengageinpotentialnegotiations,carryoutthetransactionsandsoon.Asthosemarketsbecomemature,newproductsandservices,aswellasnewpractices,arebeginningtoappear.Thereisalreadyampleevidencethatagent-basedtechnologywillbecrucialforthesedevelopments,reflectingthatbothbuyersandsellersbeingvigorouslyseektechniquestoimprovetheirperformance.ThisworksurveysseveraloftheseagentsthathelpelectroniccommercesystemsbydescribingtheirrolesinthecontextofaConsumerBuyingBehavior(CBB)model.TheCBBmodelpresentedaugmentstraditionalmarketingmodelswithconceptsfromSoftwareAgentsresearchtoaccommodateelectronicmarkets.

简介：

简介：Wepresentthesolidmodeleditdistance(SMED),apowerfulandflexibleparadigmforexploitingshapesimilaritiesamongstCADmodels.ItisdesignedtomeasurethemagnitudeofdistortionsbetweentwoCADmodelsinboundaryrepresentation(B-rep).Wegivetheformaldefinitionbyanalogywithgrapheditdistance,oneofthemostpopulargraphmatchingmethods.Toavoidtheexpensivecomputationalcostpotentiallycausedbyexactcomputation,anapproximateprocedurebasedonthealignmentoflocalstructuresetsisprovidedinaddition.Inordertoverifytheflexibility,wemakeintensiveinvestigationsonthreetypicalapplicationsinmanufacturingindustry,anddescribehowourmethodcanbeadaptedtomeetthevariousrequirements.Furthermore,amultilevelmethodisproposedtomakefurtherimprovementsofthepresentedalgorithmonbotheffectivenessandefficiency,inwhichthemodelsarehierarchicallysegmentedintotheconfigurationsoffeatures.ExperimentresultsshowthatSMEDservesasareasonablemeasurementofshapesimilarityforCADmodels,andtheproposedapproachprovidesremarkableperformanceonareal-worldCADmodeldatabase.

简介：这份报纸基于模糊规则建议修改Couzin和速度适应的Couzin结队模型。在模型，代理人在三个感觉地区通过一个fuzzy-rule-based决策计划更新他们的位置。系统的稳定性被保证。模拟证明集中概率和相对收敛尺寸两个都被改进。

简介：一个阳光灿烂的午后．当我们正在为本期的“千姬变”栏目中模特的甄选工作忙得焦头烂额的时候．摄影师Petter给我们来了一个电话：“我认识一个小姑娘学化妆的．看了你们杂志特感兴趣，要不要看看她的资料？”我们抱着多一个不多．少一个不少的心态查看了Petter发来的女孩资料。十分让人意外的是．这位学化妆的小姑娘竟然跟我们以前印象中的女性化妆师大不相同——印象中的女性化妆师虽说容颜装束都很时尚．但是大多身材并不出众，

简介：

简介：

简介：

简介：KnowledgerepresentationandreasoningisakeyissueoftheKnowledgeGrid.ThispaperproposesaKnowledgeMap(KM)modelforrepresentingandreasoningcausalknowledgeasanoverlayintheKnowledgeGrid.ItextendsFuzzyCognitiveMaps(FCMs)torepresentandreasonnotonlysimplecause-effectrelations,butalsotime-delaycausalrelations,conditionalprobabilisticcausalrelationsandsequentialrelations.ThemathematicalmodelanddynamicbehaviorsofKMarepresented.Experimentsshowthat,undercertainconditions,thedynamicbehaviorsofKMcantranslatebetweendifferentstates.Knowingthiscondition,expertscancontrolormodifytheconstructedKMwhileitsdynamicbehaviorsdonotaccordwiththeirexpectation.SimulationsandapplicationsshowthatKMismorepowerfulandnaturalthanFCMinemulatingrealworld.

简介：Parameterizationisoneofthekeyproblemsintheconstructionofacurvetointerpolateasetoforderedpoints.Weproposeanewlocalparameterizationmethodbasedonthecurvaturemodelinthispaper.Thenewmethoddeterminestheknotsbyminimizingthemaximumcurvatureofquadraticcurve.Whentheknotsbythenewmethodareusedtoconstructinterpolationcurve,theconstructedcurvehavegoodprecision.Wealsogivesomecomparisonsofthenewmethodwithexistingmethods,andourmethodcanperformbetterininterpolationerror,andtheinterpolatedcurveismorefairing.

简介：

简介：

简介：Engineeringapplicationdomainsneeddatabasemanagementsystemstosupplythemwithagoodmeansofmodeling,ahighdataaccessefficiencyandalanguageinterfacewithstrongfunctionality.Thispaperpresentsasemantichypergraphmodelbasedonrelations,inordertoexpressmany-to-manyrelationsamongobjectsbelongingtodifferentsemanticclassesinengineeringapplications.Amanagementmechanismexpressedbythemodelandthebasicdataofengineeringdatabasesaremanagedinmainmemory.Especially,differentobjectsarelinkedbydifferentkindsofsemanticsdefinedbyusers,thereforethetableswap,therecordswapandsomeunnecessaryexaminationsarreducedandtheaccessefficiencyoftheengineeringdataisincreased.Clanguageinterfacethatincludessomegenericandspecialfunctionalityisproposedforcloserconnectionwithapplicationprograms.

简介：Asemi-structureddocumenthasmorestructuredinformationcomparedtoanordinarydocument,andtherelationamongsemi-structureddocumentscanbefullyutilized.Inordertotakeadvantageofthestructureandlinkinformationinasemi-structureddocumentforbettermining,astructuredlinkvectormodel(SLVM)ispresentedinthispaper,whereavectorrepresentsadocument,andvectors'elementsaredeterminedbyterms,documentstructureandneighboringdocuments.TextminingbasedonSLVMisdescribedintheprocedureofK-meansforbriefnessandclarity:calculatingdocumentsimilarityandcalculatingclustercenter.TheclusteringbasedonSLVMperformssignificantlybetterthanthatbasedonaconventionalvectorspacemodelintheexperiments,anditsFvalueincreasesfrom0.65-0.73to0.82-0.86.

简介：Onthebasisofanalyzingsomeneuralnetworkstoragecapacityproblemsanetworkmodelcomprisinganewencodingandrecallingschemeispresented.Byusingsomelogicaloperationswhichoperateonthebinarypatternstringsduringinformationprocessingprocedurethemodelcanreachahighstoragecapacityforacertainsizeofnetworkframework.

简介：Alambdasystemwithalgebraicoperators,Lambda-plussystem,isintroduced.Aftergivingthedefinitionsofthesystem,wepresentasufficientconditionforformulatingamodelofthesystem.Finally,amodelofsuchsystemisconstructed.

简介：Amingatwaterconservancyprojectvisualization,ahidden-removalmethodofdamperspectivedrawingsisrealizedbybuildingahidden-removalmathematicalmodelforoverlappingpointslocationtosetupthehiddenrelationshipamongpointandplane,planeandplaneinspace.Onthisbasis,asanexampleofpanelrockfilldam,adamhidden-removalperspectivedrawingisgeneratedindifferentdirectionsanddifferentvisualanglesthroughadaptingVC++andOpenGLvisualizingtechnology.Theresultsshowthatthedataconstructionofthemodelissimplewhichcanovercomethedisadvantagesofconsiderableandcomplicatedcalculation.Thismethodalsoprovidesthenewmeanstodrawhidden-removalperspectivedrawingsforthoselandformsandgroundobjects.