简介:Inthispaper,thedefinitionsofBayesestimator,minimaxestimator,minimumriskequivariantestimatorandadmissibleestimatorunderthegeneralmatrixlossfunctionaregiven.Severalresultsonestimationofparameterswiththeordinarylossfunctionareextendedtothecaseofthematrixlossfunction.
简介:1.IntroductionTherehasbeenalotofliterature(see[1]--[61and[12])ontheparalleliterativemethodsforthelarge--scalesystemoflinearequationsinthesenseofmatrixmultisplittingsincethepioneeringworkofO’LearyandWhite(see[l])waspublishedin1985.Oneofthemostrecentre...
简介:Inthispaperanewnondestructivedamageidentificationmethodisintroduced.Themethodbasedonflexibilitymatrixcanbeusedtodetectandlocatestructuraldamageandevaluatetheseverityofdamageinlegsofjacketplatformsbymodalparametersofastructure.Withthemodaldataforonlythefewlowermodesinboththeintactanddamagedstates,theone-dimensionalandtwo-dimensionaldistributedcurvaturescanbeusedtoanalyzedamagelocationandtheseverity.Insteadofdirectlycomparingthecurvaturesbeforeandafterdamage,themethodhereusesmodalparametersonlyinthedamagedstructuretodetectthedamageanditconsistsofthreeparts.First,flexibilitymatrixisobtainedbyuseoftheabsolutemaximumineachcolumn.Second,becausethelegsofjacketplatformsarethepipe-likestructure,thecircumferentialflexibilitycurvaturematrixisobtainedbyuseofthecircularcurvature.Atlast,equivalentcurvatureratioisdefinedandthecurvemeaningequivalentcurvatureratioandtheseverityofdamagerelationshipforoneelementisgiventhroughthedataofdamageseverityfromtenpercenttoninetypercentbynumericalsimulation.Manyexistingdamagedetectionmethodsneedtwosteps,locatethedamagefirstlyandevaluatetheseverityofthedamage.However,themethodpresented!inthispapercanlocateandthenevaluatetheseverityofdamageatthesametime.Thenumericalanalysisresultsindicatethatthepresentmethodiseffective,usefulandonlyneedthefirstandthesecondmodedataofthestructure.
简介:InthispaperweconsideraMarkovchainmodelinanATMnetwork,whichhasbeenstudiedbyDagandStavrakakis.OnthebasisoftheiterativeformulasobtainedbyDagandStavrakakis,weobtaintheexplicitanalyticalexpressionofthetransitionprobabilitymatrix.ItisverysimpletocalculatethetransitionprobabilitiesoftheMarkovchainbytheseexpressions.Inaddition,weobtainsomeresultsaboutthestructureofthetransitionprobabilitymatrix,whicharehelpfulinnumericalcalculationandtheoreticalanalysis.
简介:GivenastablematrixA∈Cn×n,therearesomesimplesimultaneousinequalitieswhosecoefficientsrelyonA.For(?)PH=P∈Cn×n,ifPsatisfiestheseinequlities,thenPA+AHP<0.Asanapplicationofthemainresults,thestabilityofthematrixfamilyistreated.Besides,somegeometricpropertiesofthesetofalltheLyapunovmatricesofagivenstablematrixareconsidered.
简介:Theproblemofdetectingsignalwithmultipleinputmul-tipleoutput(MIMO)radarincorrelatedGaussianclutterdominatedscenariowithunknowncovariancematrixisdealtwith.Thegen-eralMIMOmodel,withwidelyseparatedsub-arraysandco-locatedantennasateachsub-array,isadopted.Firstly,thegeneralizedlikelihoodratiotest(GLRT)withknowncovariancematrixisob-tained,andthentheRaoandWalddetectorsaredevised,whichhaveprovedthattheRaoandWaldtestcoincidewithGLRTdetec-tor.Tomakethedetectorsfullyadaptive,thesecondarydatawithsignal-freewillbecollectedtoestimatethecovariance.Theper-formanceoftheproposeddetectorisanalyzed,however,itisjustancillary.Athoroughperformanceassessmentbyseveralnumer-icalexamplesisalsogiven,whichhasconsideredthesensewithco-locatedantennasconfigureoftransmittersandreceiversarray.Theresultsshowthattheperformancetheproposedadaptivede-tectorisbetterthanLJ-GLRT,andthelosscanbeacceptableincomparisontotheirnon-adaptivecounterparts.
简介:Wedevelopanumericalschemetoconstructthescattering(S)matrixforopticalmicrocavities,includingthespecialcaseswithparity-timeandothernon-Hermitiansymmetries.Thisschemeincorporatestheexplicitformofanonlocalboundarycondition,withtheincidentlightrepresentedbyaninhomogeneousterm.ThisapproachresolvestheartifactofadiscontinuousnormalderivativetypicallyfoundintheR-matrixmethod.Inaddition,weshowthat,byexcludingtheaforementionedinhomogeneousterm,thenon-HermitianHamiltonianinourapproachalsodeterminesthePeriels–Kapurstates,anditconstitutesanalternativeapproachtoderivethestandardR-matrixresultinthisbasis.Therefore,ourschemeprovidesaconvenientframeworktoexplorethebenefitsofbothapproaches.Weillustratethisboundaryvalueproblemusing1Dand2DscalarHelmholtzequations.TheeigenvaluesandpolesoftheSmatrixcalculatedusingourapproachshowgoodagreementwithresultsobtainedbyothermeans.
简介:LetAandCdenoterealn×nmatrices.Givenrealn-vectorsx1,………,xm,m≤n,andasetofnumbers(L)={λ1,λ2…λm},Wedescrbe(Ⅰ)theset()ofallrealn×nbisymmetricpositiveseidefinitematricesAsuchthatAxiisthe"best"approximatetoλixi,i=1,2,...,minFrobeniusnormand(Ⅱ)theYinset()whichminimizeFrobeniusnormof||C-Y||.AnexistencetheoremofthesolutionsforProblemⅠandProblemⅡisgivenandthegeneralexpressionofsolutionsforProblemⅠisderived.SomesufficientconditionsunderwhichProblemⅠandProblemⅡhaveanexplicitsolutionisprovided.Anumer-icalalgorithmofthesolutionforProblemⅡhasbeenpresented.
简介:
简介:阶段PVDF水晶结构和在先存在的阶段PVDF球粒以内的豌豆的结晶化上的豌豆的效果被光显微镜学OM,红外线的光谱学红外和扫描电子显微镜学SEM调查了。结果证明阶段PVDF球粒由展出高度弄弯的像滚动的形态学的lamellae组成并且在充满豌豆的混合优先地发展。与增加豌豆集中,滚动直径增加和滚动更好与对方被分开。豌豆在interspherulitic区域使第一结晶,transcrystalline层发展。随后,豌豆的transcrystalline层继续在阶段PVDF球粒以内成长,例如在在滚动之间的区域,直到在另外的豌豆transcrystalline层或球粒上侵犯。结晶化动力学结果显示在阶段PVDF的intraspherulitic区域的豌豆晶体的生长率与阶段PVDF的结晶化温度与豌豆,而是否定豌豆的内容显示出积极关联。
简介:1IutroductionManykindsofmatrix-valuedrationalinterpolationorapproximationproblemshaveappearedinrecentyears([1-7]).MotivatedbyGraves-Morris’Thiele-typevector-val-uedrationalinterpolants[6],GuChuanqingandChenZhibing[7]discussedthematrix-valuedrationalinterpolantsinThiele-typecontinuedfractionform,withmatrix-valuednu-
简介:Anewefficientcouplingrelationshipdescriptionmethodhasbeendevelopedtoprovideanautomatedandvisualizedwaytomultidisciplinarydesignoptimization(MDO)modelingandsolving.Thedisciplinaryrelationmatrix(DRM)isproposedtodescribethecouplingrelationshipaccordingtodisciplinaryinput/outputvariables,andtheMDOdefinitionhasbeenreformulatedtoadoptthenewinterfaces.Basedonthese,auniversalMDOsolvingprocedureisproposedtoestablishanautomatedandefficientwayforMDOmodelingandsolving.Throughasimpleandconvenientinitialconfiguration,MDOproblemscanbesolvedusinganyofavailableMDOarchitectureswithnofurthereffort.SeveralexamplesareusedtoverifytheproposedMDOmodelingandsolvingprocess.ResultshowsthattheDRMmethodhastheabilitytosimplifyandautomatetheMDOprocedure,andtherelatedMDOframeworkcanevaluatetheMDOproblemautomaticallyandefficiently.
简介:<正>Anadvancedsilica-dominatedmatrixtechnologyplatformanditsapplicationwereintroducedinthispaper.Throughnovelproprietaryprocessing,theadvancedsilica-dominatedmatrixwaspreparedfromtheuniquesilicasolandpseudo-boehmite.Thematrixporesizedistributionandaciditycanbetailoredtoprovideoptimalcoke-selectivityandbottoms-crackingabilitymatchedtofeedstockcharacteristics.AnovelFCCcatalyst(RSC-2006)forreducingcokeandslurryyieldswasdevelopedbasedonthesilica-dominatedmatrixtechnologyplatform.TheresultsofcommercialapplicationofthesaidtechnologyindicatedthattheRSC-2006catalystexhibitedexcellentperformanceonbottomscracking,cokeselectivityandhighvalueproductsyields.