International Journal of Radiation OncologyBiologyPhysicsWord文档格式.docx
《International Journal of Radiation OncologyBiologyPhysicsWord文档格式.docx》由会员分享,可在线阅读,更多相关《International Journal of Radiation OncologyBiologyPhysicsWord文档格式.docx(26页珍藏版)》请在冰豆网上搜索。
Aprocedurefortheoptimizationofwind-excitedstructuresisproposed.Itisbasedonthesimulatedannealingalgorithmcombinedwiththedynamicanalysisoftheresponseeitherinthefrequencyorinthetimedomain.Whenthestep-by-stepdynamicanalysisisused,itcanhandlethecaseofflexiblestructures,likemastsandlatticetowers,whosegeometricnon-lineareffectscannotbeneglected.Theprocedureallowsformultiplevariablesandobjectives.Theproposedmethodisusedtooptimizetheconfigurationofacable-stayedmastsubjectedtoturbulentwindloading.Theresultsshowedthatthealgorithmisreasonablyindependentofthefirstguessconfigurationandiseffectiveinavoidinglocalminima.Itswidefieldofapplicabilityandeaseofimplementationmaketheproposedalgorithmapowerfuldesigntoolforstructuralengineers.
ArticleOutline
1.Introduction
2.Theoptimizationtechnique
2.1.TheSimulatedAnnealingalgorithm
2.2.Themultipleobjectivefunction
2.3.Implementationoftheprocedure
3.Windloadingandstructuralresponse
3.1.Windloadmodeling
3.2.Frequencydomainanalysis
3.3.Timedomainanalysis
4.Numericalapplication
4.1.Descriptionofthecasestudy
4.2.Theoptimizationproblem
4.3.Results
5.Concludingremarks
Acknowledgements
References
Purchase
$31.50
687
UFOme:
Anontologymappingsystemwithstrategypredictioncapabilities
Data&
KnowledgeEngineering,Volume69,Issue5,May2010,Pages444-471
GiuseppePirró
DomenicoTalia
Ontologymapping,ormatching,aimsatidentifyingcorrespondencesamongentitiesindifferentontologies.Severalstrandsofresearchcomeupwithalgorithmsoftencombiningmultiplemappingstrategiestoimprovethemappingaccuracy.However,fewapproacheshavesystematicallyinvestigatedtherequirementsofamappingsystembothfromthefunctional(i.e.,thefeaturesthatarerequired)anduserpointofview(i.e.,howtheusercanexploitthesefeatures).Thispaperpresentsanontologymappingsoftwareframeworkthathasbeendesignedandimplementedtohelpusers(bothexpertandnon-expert)indesigningand/orexploitingcomprehensivemappingsystems.Itisbasedonalibraryofmappingmodulesimplementingfunctionssuchasdiscoveringmappingsorevaluatingmappingstrategies.Inparticular,thestrategypredictormoduleofthedesignedframework,foreachspecificmappingtask,can“predict”mappingmodulestobeexploitedandparametervalues(e.g.,weightsandthresholds).Theimplementedsystem,calledUFOme,assistsusersduringthevariousphasesofamappingtaskexecutionbyprovidingauserfriendlyontologymappingenvironment.TheUFOmeimplementationanditspredictioncapabilitiesandaccuracywereevaluatedontheOntologyAlignmentEvaluationInitiativetestswithencouragingresults.
2.Definitionsandproblemstatement
2.1.Ontology
2.1.1.Ontologicalinformation
2.2.Ontologymapping:
motivationandexample
2.3.Ontologymappingrepresentation
2.4.Ontologymappingsystem
2.5.Mappingstrategy
2.6.Similarityfunction
3.Relatedwork
3.1.Usersupportforontologymapping
3.2.Ontologymappingsystemsbasedonmultiplestrategies
3.3.Evaluatingmappingstrategies
3.4.Automaticparameterstuning
3.5.Requirementsofanontologymappingframework
3.6.Whatamappingsystemshouldfeature?
4.Alibraryofmodulesforontologymappingsystems
4.1.Visualization
4.2.Matching
4.2.1.TheLuceneontologymatcher(LOM)
4.2.2.Thewordnetmatcher(WM)
4.2.3.Thestringmatcher(SM)
4.2.4.Thestructuralontologymatcher(SOM)
4.3.Combination
4.4.Strategyprediction:
thestrategypredictormodule
4.4.1.Lexicalaffinitycoefficient
4.4.2.Structuralaffinitycoefficient
4.4.3.Exploitingaffinitycoefficients
4.4.4.Ondeterminingtheoptimalthresholdvalues
4.4.5.Ondeterminingoptimalweightvalues
4.5.Evaluation
5.UFOme:
Acomprehensiveontologymappingsystem
5.1.Areferenceontologymappingframeworkarchitecture
5.2.TheUFOmesystem
5.2.1.TheUFOmetwo-layerarchitecture
5.2.2.Phase1:
Designing
5.2.3.Phase2:
Running
5.2.4.Phase3:
Evaluation
5.2.5.Remark
6.Experimentalevaluation
6.1.Thedatasetandtheevaluationmethodology
6.2.Evaluationofthestrategypredictormodule
6.2.1.Parametertuning
6.3.Effectofthethresholdonlinguisticmappingstrategies
6.4.Effectofthethresholdonstructuralontologymapping
6.5.Evaluatingautomaticweightsassignment
6.5.1.Evaluatingtheeffectsofmappingmodulesselection
6.5.2.ComparingUFOmewithothersystems
7.Discussion
8.Futurework
Vitae
688
Anefficientdiagnostictechniquefordistributionsystemsbasedonunderfaultvoltagesandcurrents
ElectricPowerSystemsResearch,Volume80,Issue10,October2010,Pages1205-1214
A.Campoccia,M.L.DiSilvestre,I.Incontrera,E.RivaSanseverino,G.Spoto
Servicecontinuityisoneofthemajoraspectsinthedefinitionofthequalityoftheelectricalenergy,forthisreasontheresearchinthefieldoffaultsdiagnosticfordistributionsystemsisspreadingevermore.Moreovertheincreasinginterestaroundmoderndistributionsystemsautomationformanagementpurposesgivesfaultsdiagnosticsmoretoolstodetectoutagespreciselyandinshorttimes.Inthispaper,theapplicabilityofanefficientfaultlocationandcharacterizationmethodologywithinacentralizedmonitoringsystemisdiscussed.Themethodology,appropriateforanykindoffault,isbasedontheuseoftheanalyticalmodelofthenetworklinesandusesthefundamentalcomponentsrmsvaluestakenfromthetransientmeasuresoflinecurrentsandvoltagesattheMV/LVsubstations.Thefaultlocationandidentificationalgorithm,proposedbytheauthorsandsuitablyrestated,hasbeenimplementedonamicroprocessor-baseddevicethatcanbeinstalledateachMV/LVsubstation.Thespeedandprecisionofthealgorithmhavebeentestedagainsttheerrorsderivingfromthefundamentalextractionwithintheprescribedfaultclearingtimesandagainsttheinherentprecisionoftheelectronicdeviceusedforcomputation.ThetestshavebeencarriedoutusingMatlabSimulinkforsimulatingthefaultedsystem.
2.Thediagnosticsystem
3.Circuitmodelofonelinespanfordiagnosys
3.1.Unfaultedcondition
3.2.Faultedcondition
4.Faultidentificationandcharacterization
4.1.Solutionalgorithm
4.2.Phaseassessment
5.Compliancetoexistingsystems
6.Applications
6.1.Precisionofthediagnosticalgorithm
6.1.1.Singlephasetoearthfault
6.1.2.Threephasetogroundfault
6.2.Calculationtimes
7.Conclusions
AppendixA.Microprocessorimplementation
$41.95
689
On-linehandwrittendigitrecognitionbasedontrajectoryandvelocitymodeling
PatternRecognitionLetters,Volume29,Issue5,1April2008,Pages580-594
MonjiKherallah,LobnaHaddad,AdelM.Alimi,AmarMitiche
Thehandwritingisoneofthemostfamiliarcommunicationmedia.Penbasedinterfacecombinedwithautomatichandwritingrecognitionoffersaveryeasyandnaturalinputmethod.Thehandwrittensignalison-linecollectedviaadigitizingdevice,anditisclassifiedasonepre-specifiedsetofcharacters.Themaintechniquesappliedinourworkincludetwofieldsofresearch.Thefirstoneconsistsofthemodelingsystemofhandwriting.Inthisarea,wedevelopedanovelmethodofthehandwrittentrajectorymodelingbasedonellipticandBetarepresentation.ThesecondpartofourworkshowstheimplementationofaclassifierconsistingoftheMulti-LayersPerceptionofNeuralNetworks(MLPNN)developedinafuzzyconcept.ThetrainingprocessoftherecognitionsystemisbasedonanassociationoftheSelfOrganizationMaps(SOM)withFuzzyK-NearestNeighborAlgorithms(FKNNA).Totesttheperformanceofoursystemwebuild30,000Arabicdigits.Theglobalrecognitionrateobtainedbyourrecognitionsystemisabout95.08%.
2.TrajectorymodelingbyBeta–ellipticalrepresentation
2.1.Betavelocitymodeling
2.2.Ellipticaltrajectorymodeling
2.3.CombinationbetweenBetaandellipticalmodels
3.On-linerecognitionofhandwrittendigits
3.1.Pre-processingsystem
3.2.Realclassdetectionbyself-organizingmap
3.3.MembershipassignmentofthetrainingdatasettorealclassesbytheFuzzyK-NearestNeighborAlgorithm
3.4.ClassificationbytheMLPNNsystem
3.5.Digitdatabaseformulation
4.Experimentalresultsanddiscussions
5.Conclusion
690
EDICAMfastvideodiagnosticinstallationontheCOMPASStokamak
FusionEngineeringandDesign,Volume85,Issues3-