A Comparative Study on Optimization in Structural Acoustics [Elektronische Ressource] / Mostafa Ranjbar. Gutachter: Hans-Jürgen Hardtke ; Steffen Marburg ; Ariosto Bretanha Jorge. Betreuer: Hans-Jürgen Hardtke

Description

GutacEineerteidigung:vProf.ergleicUnivhendeh.c.StudieEinreiczurTOptimierungoninProf.derProf.StrukturakustikJorge(ATComparativesenehniscStudyDresdenonMostafaOptimizationterinhabil.StructuralHardtkASteencoustics)AriostoDagIung:SderSwEderRecThenAersit?tTvIM.Sc.(Eng.)ORanjbarNhzur:ErlangungDr.-Ing.desProf.akHans-J?rgenademiscehenDr.-Ing.GradesMarburgDoktor-IngenieurDr.(Dr.-Ing.)BretanhavTorgelegtderderhF09/09/2009akult?tagf?rVMasc25/03/2011hinenacousticAbstractforThisddissertationofpresenobtsandanofexhaustiveeascomparativNewtonesearcstudyalueonhoptimizationevinasstructuralwacoustics.pAicomfeasiblebinationmethoofoinads,commerciallymethoaavanailableTheniteelelemenastoinsoftiswtheareedpactkcase,ageandandativadditionale.g.,user-writtenquadraticprogramsmemoryisoundusedasymptotes,toandmoThedifyrithms,theareshapuseeeofisamethostructure.anThisrepisvdoneanditer-optimizationativmethoelysloandandwithoutaremands.ualtheindstervIfendstion(fasttothenacresultshievcomputationeensignicantthimprosecondvofemenetsdofsequenthemethoobd,jectivyden-Fletceforfunction.problems,ThemooptimizationmpromethocesstrolledconmethotinoptimizationuesgeneticautomaticallysearcunulatedtilativtheandpredenedexactmaximobumFnapproacumwbcomparisonerAofandfunctionofevalgorithmaluationsinisofreac(ahed.

Sujets

• Physik

Informations

• Publié par: technische_universitat_dresden
• Publié le: 01 janvier 2011
• Nombre de lectures: 12
• Poids de l'ouvrage: 2 Mo

Extrait

Einevergleicerteidigung:GutacProf.hendeUnivStudieh.c.zurEinreicOptimierungTinonderProf.StrukturakustikProf.(AJorgeComparativTeesenStudyhnisconDresdenOptimizationMostafainterStructuralhabil.AHardtkcoustics)SteenDAriostoIagSung:SderEwRderTecAhenTersit?tIvOM.Sc.(Eng.)NRanjbarzurhErlangung:desDr.-Ing.akProf.ademiscHans-J?rgenheneGradesDr.-Ing.Doktor-IngenieurMarburg(Dr.-Ing.)Dr.vBretanhaorgelegtTderderFhakult?t09/09/2009f?ragMascVhinen25/03/2011AbstractThisdissertationpresentheanalysisoftsmoanjectivexhaustivsimeoptimizationcomparativstructural-acousticeastudymethoonhoptimizationofinoptimizationstructuralaluated.acoustics.theAerfulcomeriobinationtheofdirections,adcommerciallytsae.g.,vdsailablestatisticalnitetageselemenratetofsoftfast,wtsareexppactotalkwithagemoreandtheyadditionaltheuser-writtenesprogramsmethoisprogrammingusedBrotoconstrainedmomid-rangedifycontheexactshaptabuederivoftheafunction.structure.folloThisds.istagesdoneortediter-ergenceativtheelymethoanddswithoutw.manmid-rangeualininFinallytervuseencantionthetobacandhievtheyeinsignicantime.tbimproforvybridemenderivtstheoffunction,theofobtialjectivd,elimitedfunction.her-Goldfarb-ShannoTheboptimizationmethoprovingcessulti-pcondtinrandomuesd.automaticallymethounalgo-tilhtheannealing,predenede-freemaximtheyumvnjectivumurthermore,bheredofoffunctiondvevdisadvaluationseacisarereacdetails.hed.conThemeasuredesigneed)vlevariablesharearetheoptimizationstructure'sclassiedloandcaldgeometryasymptotesmoulti-pdicationdvducedaluesmethoatitselectedthatsurfaceeectivkmethoey-preduceointime.ts.oThemethoobequippjectiveestructuralofds,thepresenoptimizationoptimizationincludesshortertheofminimizationthisofevtheconsideredroreplacemenotcomplexmeanulti-stagessquarealgorithms.levsecondelativofofstructureobbeornee.g.,soundd(afeasiblegeneralsequenmeasurequadraticofmethotheNewtonvibrationald,sensitivitmemoryyyden-Fletcofmethoaforstructure).oundInproblems,addition,dthemostructuralasymptotes,massmremainsoinconstanmethotandandtrolledthesearcallomethowTheableoptimizationrangesds,ofgeneticdesignrithms,vsearcariableandvulatedaluesareareativrestrictedmethobandyuseprescribexactedalueuppobereandFloawapproacerislimits.wTheforoptimizationcomparisonpromethocedureAisantestedandonantheofnitehelemenalgorithmtrepmoindelTheofofavrectangular(aplateofmadespofandsteel.robustnessTelweacelvoptimizationeddierenevtSomeoptimizationmethomethoaredsasaremediumtestedsloagainstMethoeacofhvingothers.andThesemmethooindsmethoareareconsideredtroeitherasasfastestapprods.ximate,orisexact.eriencedThetheapproofximateeoptimizationanalysismethodsdsdrasticallyusetheeitheroptimizationanIfappropximatedwvoptimizationaluedsofecomeobedjectiveective(fastfunction,reliable)e.g.,acoustichmethoybridthendesigncanoftexpdesirableerimenresultstsaandphdybridcomputationneuralInnetcase,wcanorks,enoretheasapprosuitableximatedtvthealuesandofmthehrstoptimizationandiZusammenfassungDievzuonworliegendedeDissertationendenpr?senvtiertdieeinede,vTergleicstatistischendeundStudieWzurtenOptimierungzwinderdervStrukturakustik.sindEineDieseKVomor-bination(einauserdeneinemalskendungommerziellendieFinite-Elemenneuronalete-SoftertewZielfunktion;are-hPquadratiscakerfahren,etMid-Range-Multi-Punkt-MethoundDiedergenetiscimundRahmenden.derertArbOptimierungsveiterwenteiletKwicOptimierungsgesckeinzelneneltenDiezus?tzlicundhener-ProgrammedasswirdstrukturakustiscvVerwRecen-hdet,oumeisendieerstenGeometrieAbleitungeneinerB.meczul?ssigenhaniscMethohenquenStrukturProgrammierung,zuGSoptimieren.deDerhenautomatiscundheZufalls-Op-VtimierungsprozessB.wirdAlgorithmsoSearclangeulateddurcableitungsfreiehgef?hrt,vbisgenaueneineZielfunktion.vhorgegebwirdeneAnsatzmaximaleDieAnzahlNacanMethoAuswdiskutiert.ertungenvderf?rZielfunktionwindigkerreicRobustheithdentewwird.deEnhentdewurfsvteariablenidensindwirddieVlokezienalenAnaly-Geometrie?nderungendrastiscanderausgewb?hltenf?hrt.Obunder?cybridehen-Netze;punkten.derDasn?herungswZielWderderOptimierungundumfassteitendiederMinimierungz.desMethoEektivwderertesRicdertungen,Scdehall-se-leistung.tiellenDerhenzul?ssigeNewton-MethoBereicBFhVderMethoEnderter?nderlicwurfsvAsymptoten,ariablendewirdkbtrollierteegrenzt.iteration.DasexaktenOptimierungsverfahrener-z.fahrenderwirdheanus,eineraburechhSimtecAnnealing.kigensindPlatteMethoausSieStahlerwerprobt.denZwW?lfdervZumerscergleichiedenederOptimierungsverfahrenerfahreneinwhererdenvmiteinanderendet.vVerglicundhen.hDiesederVdener-erdenfahrenDiesindonenergenzratetMa?wdieederhexakteit)odiederderapproOptimierungsmethoximativ.wDiebapproertet.ximativMethoenderMethoer?nderlicdenAsymptotenvMid-Range-Multi-Punkt-MethoerwwurdenendenezienenOptimierungsvtfahrenwtiziert.edereiterhineinengezeigt,n?herungswdieeisenerwWeinererttenderhenZielfunktion;semethodez.einerB.henHybrid-Designerringerungvf?ronOptimierungExpen?tigteneri-henzeitmeniiAcknoceherwledegmenthetsectivIwhowofouldmemlikisegoals.totoleratingacuxknousedwledgeTtheFP6suppProf.ortinofythismenwfororkbbouldyknothefromInstitutehniscofISolidtheMecAhanicsunitfromProgramme.theandFofaculttoyofofthankMecandhanicalmEngineeringenabledoflothemTyecthehnicalofUnivestersitduringyAltixofPCDresden,f?randofbDresdenythemvisityMarccolleaguesUnivfromundertheI3-CT-2003-506079)indeptheendenAteangrouperyfortheStructuralTU/eAforcousticsIOptimizationyResearcB?shes.inIofwishpleasuretoteacthankstudiesmdaresyareacademicforsupsuppervisormProf.expressDr.-Ing.mProf.uncomplainingh.c.wHans-J?rgenThanksHardtktoeyandydirectortheofFinallyInstitutexpressf?revFwithestkeducations.?pthatermecandhanikw(IFKM),arm-PhobforZengivingundmehnenaTPhDUnivscwholarshipcarryforwtheadurationtheofebruaryOctob2007erv2003ytohnologyMarcHPC-EURhject2007.ork,IortameanvheryunderthankfulthetohmamytoscienersticforsuptrolervisorbProf.HenkDr.-Ing.assistancesSteenyMarburgouldforexpresstheestconceptualJoacformgivingulationaluableofbthisprothesisdissertation.andahismevmaluableduringandandgreatthescienarbiatticersities,guidance,.forduesharingparenthetinknoandwledgewhicintothisideaseldwishofyresearcehwifeandsteadyfororttheresearcsteadyanddiscussionatoftoforthcomingonderfulwaork.duringIstudyameautifulvDresden,erymethankfulprotoyDipl.-Ing.IDenneyyFectsritzeoforedfruitfulydiscussionsmabacoutwledgethisSGIw3700-Merkurork.LinINetworkouldFlikosethetotrumthankInformationsdienstemHoyhleistungsreccolleagues,(ZIH)e.g.,theDr.ecrer.henat.ersit?tElk(TUD)eereJunctokoutert,computationalDr.-Ing.ork.JoachadhimfruitfulGier,inDipl.-Ing.durationHans-JoacFhimtoBeer,hPfrometraEindhoWenesz,ersitDr.-Ing.ofStefanecSc(TU/e)hneider,theDr.-Ing.OPMicprohael(RIScframewheer,withDr.-Ing.suppIrwofanEuropto,CommDr.y-ResearcZhirongInfrastructureWctionang,theDipl.-Ing.StructuringAnjaEuropJablonski,ResearcHergartAreaWIolf-HoppvethankfulandtheDipl.-Ing.bRobofertinstituteAnderssohn,DynamicswhoConwinenheadingtywithDr.meNijmeijerthroughtheirtheseduringymearsvisit.andwcreatedlikthetonicematmospherebduringgratitudeandDr.-Ing.outhimofforthemewvorkassistancetime.theIeginningwwritingancesstthistoItthankalsoProf.greatDr.forrer.tonat.allWyolfganghersNagelmwhoundergraduategagraduatevateShirazmeTtheMopUnivossibilitrespyelytoThankswalsoorktoinythetsCenconteruousfortalInformationnancialandort,HighhPmeerformancepursueComputingyofandtheITtoecmhnicalgreatUnivversittoyyofOzraDresdenherfromandMasuppyoftoyAugusthofork2006.forIcareamhome.indebtedalsotomthewscienson,ticPuyco-wwhoorkornersmofPh.D.ZIH,inespbeciallycittoofDr.forrer.withnat.duringMatthiaswritingM?llercessandmMatthiasthesis.Lieb,erwforlikdelivtoeringmexcellenbtrespfeedbactokserybondymhelpymewanork.formatIofwishytoiiiDEDICATEDTOTHE(1957-2007)MOTHERREZAIEESPIRITVEDOFMANIJHEHMYDEHSHIBI,BELOivContentsAbstract.of.i.Zusammenfassung.ii.A.cEquationknoOptimizationwledegmen.ts.iii.Listareasof.Sym.b.ols.andOptimizationA.cron3.2.2ymsdsixLo1ximationIn.tro.duction.1.1.1.Bac.kgroundStructuraland.Motiv25ation..Algorithms.........................36.....Op.......23...........2.11A1.2.Literature.ReviewMean.....Optimization.Asp.............cal.......ds.....3.3.........Metho.....36.......ximation......2.1.2.122OptimizationResearcin.Structural.A.coustics........Outline...................23.coustics.undamen.Mac..2.1.2.2.Numerical.Optimization..Ro.Lev.Borne.......26.Appro.3.1.of.......29.ds...............3.2.1.Metho..12.1.2.3.Structural.Optimization....Optimization.............ximation...............35.Appro.............Global14ds1.2.4.T.op.ology.Optimization.andMid-RangeFdsully.Stressed.Design....36.........1.4.3.en.h....15.1.2.5.Numerical.A.coustics.in.General............1.5..............................15.1.2.6.A.ctiv.e.Con2trolA.25.F.tal.of.hine.coustics...................2.2.ot.Square.el.Structure.Sound...............3.and17ximation1.329HybridGeneralRobustectsShapNumericaleAlgorithmsOptimization............3.2.Metho...............................3218Lo1.4OptimizationScopdse.and.Ob.jectiv.es.of.this.Study........32.Global.Metho.......................34.Appro.Metho......20.1.4.1.Issues.that.are.Ob.jectiv.es..........3.3.1.cal.ximation.ds.......................3.3.2.Appro.Metho....20.1.4.2.Issues.that.are.not.Ob.jectiv.es..3.3.3.Appro.Metho.....................v4OptimizationProMetho84.cedure.39.4.1MethoOptimization.ProoncedureSearcin.General..........................Results.............Asymptotes......39.4.1.189Calculation.of.Ob.jectiv.e.ObTjectiv.e.F.unctionResultsand.Constrain.ts...........41F4.1.2.Start.oforksOptimization.Pro.cedureA.1...d.....85.......Sim...........Neural.......ulated....41.4.1.3MethoEv.aluation.of.Motrolleddied.Design....Metho...6.4.Design.6.4.1.......of.......75.........77.79....41.4.2.F7.2orm.ulation.of.the.Optimization80Proofcessd.....A.2.Directions.....Sequen.....L-BF.......for.....88.d....42Genetic4.3.Con.v91ergence.Criteria......6.2.8.w.............6.2.9.Metho...........68.Searc...............6.2.11.Searc......43.4.4.Con6.3trolGeneticP.arameter.Setting....of.of.....History.............6.5.Metho...............P............43.4.5.Calculation.ofConclusionsObwjectivConclusionse.F.unction..............uture...............