简介:Shipcollisionaccidentsarerareeventsbutposehugethreattohumanlives,assets,andtheenvironment.Manyresearchershavesoughtforeffectivemodelsthatcomputeshipstochasticresponseduringcollisionsbyconsideringthevariabilityofshipcollisionscenarioparameters.However,theexistingmodelswerelimitedbythecapabilityofthecollisioncomputationalmodelsanddidnotcompletelycapturecollisionscenario,andmaterialandgeometricuncertainties.Inthispaper,anovelframeworktoperformancecharacterisationofshipsincollisioninvolvingavarietyofstrikingshipsisdeveloped,bycharacterisingthestructuralconsequenceswithefficientresponsemodels.Adouble-hulloilcarrierischosenasthestruckshiptodemonstratetheapplicabilityoftheproposedframework.Responsesurfacetechniquesareemployedtogeneratethemostprobableinputdesignsetswhichareusedtosampleanautomatedfiniteelementtooltocomputethechosenstructuralconsequences.Theresultingpredictor-responserelationshipsarefittedwithsuitablesurrogatemodelstoprobabilisticallycharacterisethestruckshipdamageundercollisions.Asdemonstratedinthispaper,suchmodelsareextremelyusefultoreducethecomputationalcomplexityinobtainingprobabilisticdesignmeasuresforshipstructures.Theproposedprobabilisticapproachisalsocombinedwithavailablecollisionfrequencymodelsfromliteraturetodemonstratetherisktolerancecomputations.
简介:ADRNN(diagonalrecurrentneuralnetwork)anditsRPE(recurrentpredictionerror)learningalgorithmareproposedinthispaper.UsingofthesimplestructureofDRNNcanreducethecapacityofcalculation.TheprincipleofRPElearningalgorithmistoadjustweightsalongthedirectionofGauss-Newton.Meanwhile,itisunnecessarytocalculatethesecondlocalderivativeandtheinversematrixes,whoseunbiasednessisproved.Withapplicationtotheextremelyshorttimepredictionoflargeshippitch,satisfactoryresultsareobtained.Predictioneffectofthisalgorithmiscomparedwiththatofauto-regressionandperiodicaldiagrammethod,andcomparisonresultsshowthattheproposedalgorithmisfeasible.
简介:Inthispaperweshowthcmethodofenergyinpartwithwhichwecangetthemodelofrandomwave,andpredicttherollmotionofunstabilizedshipandstabilizedshipusingthewavemodel.Thecontrolparametersoffinstabilizeraredeterminedaccordingtotheperformanceindex.Thesimulationofthesystemisalsomadeinthispaper.Thecomparisonofthesimulationwithrealshipindicatesthatthemethodcanbeusedinthepredictionofrollmotionofastabilizedshipinrandomwave.
简介:SHIPHYDRODYNAMICSBASICRESEARCHINTHEUNITEDSTATES¥EdwinP.Rood(MechanicsandEngergyConversion(333),OfficeofNavalResearchArlington...
简介:Inthisarticle,anewacoustictesttechniqueusingtowedmodelwasintroducedtostudyflownoisecausedbyasurfaceship.Theprojectofmodeltestwasbeproperlydesignedforacousticsignalcollectingandwiththehelpofappropriatedataprocessingmethoddifferentkindsofacousticsourcescouldbesuccessfullyidentified.Alotofworkaboutfuidnoisecouldbecarriedonwiththetowedmodel,andthenoisecorrespondingtolowfrequencywhichisespeciallyinterestedforitslongdistanceradiatingwithsmallattenuationcouldalsobestudiedinthisway.
简介:Hydrodynamicoptimizationofthehullformscanberealizedthroughtheimplementationandintegrationofcomputationaltoolsthatconsistofahydrodynamicmodule,ahullsurfacerepresentationandmodificationmodule,andanoptimizationmodule.Inthepresentpaper,anewbulbousbowgenerationandmodificationtechniquehasbeendevelopedandintegratedintothehullsurfacerepresentationandmodificationmodule.Aradialbasisfunctionbasedsurrogatemodelisdevelopedtoapproximatetheobjectivefunctionsandreducethecomputingcost.Amulti-objectiveartificialbeecolonyoptimizationalgorithmisimplementedandintegratedintotheoptimizationmodule.Toillustratetheintegratedhydrodynamicoptimizationtools,acargoshipisoptimizedforreduceddrag.Theoptimalhullformsobtainedarethenvalidatedcomputationallyandexperimentally.Validationresultsshowthatthepresenttoolscanbeusedefficientlyandeffectivelyinthesimulationbaseddesignofthehullformsforreduceddrag.
简介:Thesimulationofthewholeship-bridgecollisionprocesscanbeeffectivelycarriedoutbynonlineardynamicfiniteelementmethod.Basedonthesimpledescriptionofthetheory,ascenarioofa40000DWToiltankercollidingwithabridgeacrosstheYangtzeRiverisdesignedforsimulation.Thetechnologyofstructuremodelingandthedeterminationofrelatedparametersareintroduced.Thedeformationofthebulbbow,thehistoryofcollisionforcechange,theexchangeofcollisionenergyandthestressdistributionofthebridgepieraredescribedindetail,whichareofgreatvaluetobridgede-signandbridgepierdamageestimation,Somemechanicalcharactersintheprocessofship-bridgecollisionaredescribed.Moreaccurateresultscanbeproducedbyfiniteelementmethodthanthatbyempiricalformulasandsimplifiedanalyticalmethods.
简介:Theshipdesignprocessmodelisthebasisfordevelopingtheshipcollaborativedesignsystemundernetworkenvironment.Accordingtothecharacteristicsoftheshipdesign,amethodfordividingtheshipdesignprocessintothreelayersispatforward,thatisprojectlayer,designtasklayeranddesignactivitylayer,thentheformalizeddefinitionsoftheshipdesignprocessmodel,thedecomposingprinciplesoftheshipdesignprocessandthearchitectureoftheshipcollaborativedesign(SDPM)systemarepresented.Thismethodsimplifiestheactivitynetwork,makestheoptimizationandadjustmentofthedesignplanconvenientandalsomakesthedesignprocesseasiertocontrolandchange,atlastthearchitectureoftheshipcollaborativedesignsystemisdiscussed.
简介:Atheoreticalmodelwasusedtoinvestigatetheinfluencesofenvironmentalandtopographicparametersontheatmosphericthree-layerinternalshipwaves.TheresultsshowthatboththewavelengthandthewedgeangledecreasewithanincreaseintheScorerparameter,andtherateofchangesinthewavelengthandwedgeanglearealsoalike.Theresultsalsoshowthatthewedgeangledecreaseswithanincreaseinthewidthofmountain,butthewavelengthvarieslittlewithit.Itissuggestedthatthewedgeangleisdeterminedbytheratioofthewavelengthtothescaleofthemountain.Besides,numericalsensitivityexperimentswereperformedtotesttheformernumericalexperiments.Thesimulatedresultsareconsistentwiththeanalyticalresults.
简介:Onthebasisofenvelopespectrumanalysisofship-radiatednoise,thefeaturesandtheextractingmethodsarediscussed,andthenanewmethodofdynamicrecognitionhasbeenproposed.Usingthismethod,accuraterecognitionratefordifferenttypesofshiptargetscanbeeffectivelyimproved,andsomeoftheimportantparametersoftarget'smovementcanbeobtainedatthesametime.