时域积分方程时间步进算法及其快速算法学士学位论文撰写范例英文.docx
《时域积分方程时间步进算法及其快速算法学士学位论文撰写范例英文.docx》由会员分享,可在线阅读,更多相关《时域积分方程时间步进算法及其快速算法学士学位论文撰写范例英文.docx(14页珍藏版)》请在冰点文库上搜索。
时域积分方程时间步进算法及其快速算法学士学位论文撰写范例英文
摘要
本文基于电磁理论中的Calderόn关系与Calderόn恒等式所揭示的不同积分算子之间的关系,系统地研究了Calderόn预条件技术及其在计算电磁学积分方程方法中的应用。
研究内容全面覆盖了求解理想导电体目标和均匀或分层均匀介质目标电磁散射与辐射问题的积分方程中的Calderόn预条件技术。
在导体积分方程方面,研究了电场积分方程在中频,低频,以及高频区的Calderόn预处理方法。
在介质积分方程方面,则研究了PMCHWT积分方程的Calderόn预处理方法,和N-Müller积分方程的Calderόn技术。
本文对金属问题中的第二类Fredholm积分方程和介质问题中的第二类Fredholm积分方的精度改善进行了深入详尽的研究。
……
关键词:
电磁散射,面积分方程方法,Calderόn预条件方法,数值计算精度,第二类Fredholm积分方程
ABSTRACT
RevealedbytheCalderόnrelationandtheCalderόnidentitiesinelectromagnetictheory,thepropertiesandrelationofdifferentintegraloperatorsinthecomputationalelectromagnetics(CEM)areutilizedtoconstructtheCalderόnpreconditioningtechniques,whichareappliedintheintegral-equation-basedmethodsinthisthesis.AthoroughandsystematicresearchhasbeenaccomplishedtocovertheCalderόnpreconditioningtechniquesfortheperfectelectricconductor(PEC)andthedielectriccases.ForthePECcase,theCalderόnpreconditioningfortheelectric-fieldintegralequation(EFIE)atmid,low,andhighfrequenciesareconstructedandstudied.Forthedielectriccases,theCalderόnpreconditioningforthePoggio-Miller-Chang-Harrington-
Wu-Tsai(PMCHWT)integralequationareinvestigated,andtheCalderόntechniquefortheN-Müllerintegralequationisdeveloped.Moreover,theaccuracyimprovingtechniqueforthesecond-kindFredholmintegralequationforbothPECanddielectriccasesisalsostudiedinthisthesis.
…
Keywords:
Electromagneticscatteringandradiation,surface-integral-equation-basedMethods,Calderόnpreconditioningmethods,numericalaccuracy,Fred-
holmintegralequationsofthesecondkind
Contents
Chapter1Introduction1
1.1ResearchBackgroundandSignificance1
1.2StateofArts1
1.3ContentsandInnovationsoftheThesis2
1.4OutlineoftheThesis2
Chapter2TheoreticalBasics3
2.1IntegralEquationsinElectromagnetics3
2.2270MHzPlanWaveExcitation3
2.3TheSolutionofIntegralEquationsinElectromagnetics4
2.3.1GeneralPrincipleoftheMethodofMoments4
2.3.2GeometricalModelingandDiscretizationofObject4
2.3.2.1PlanarTriangularModel4
2.3.2.2CurvilinearTriangularModel4
2.3.3TheChoiceofBasisFunctions5
2.3.3.1PlanarRWGBasisFunctions6
2.3.3.2CurvilinearRWGBasisFunctions6
2.3.4TheSolutionofMatrixEquations6
2.3.4.1DirectAlgorithms6
2.3.4.2IterativeAlgorithms6
2.4Conclusion6
Chapter3CalderόnPreconditioneratMidFrequencies7
3.1Introduction7
3.2CalderόnRelationandCalderόnIdentities7
3.3CalderόnPreconditioneratMidFrequencies7
3.4NumericalExamples7
3.5Conclusion7
Chapter4CalderόnPreconditioningTechniqueforN-Müller8
4.1Introduction8
4.2N-MüllerIntegralEquations8
4.3TheDerivationofN-MüllerEquations8
4.4TheDiscretizationofN-MüllerEquations8
4.5NumericalExamples8
4.6Conclusion8
Chapter5Conclusions9
5.1ConcludingRemarks9
5.2FutureWork9
Acknowledgements10
References11
OriginofForeignLanguageMaterials12
TranslationofForeignLanguageMaterials13
Chapter1Introduction
1.1ResearchBackgroundandSignificance
Integral-equation-basednumericalmethodscombinedwithfastalgorithmsarecapableofsolvingelectromagneticproblemsofcomplexstructuresandmaterialpropertieswithagoodaccuracyandahighefficiency.Theyarewidelyusedinavarietyofengineeringapplications,suchastheefficientanalysisofthreedimensionalradarscatteringproblems,thesimulationoftheinputimpedanceandtheradiationpropertiesofantennasystems,thecalculationoftheinputresponseandthetransmissionefficiencyofmicrowavecircuits,theevaluationoftheelectromagneticinterference(EMI)betweencomplexelectromagneticsystems,andthecomputeraidedelectromagneticcompatibility(EMC)designs.Theversatility,capability,accuracyandefficiencyoftheintegral-equation-basedmethodshavemadethemanimportantandcosteffectiveapproachintheanalysisanddesignofelectromagneticproblemsandapplications.
…
1.2StateofArts
Fromthe1960s,thenumericalmethodsofelectromagneticanalysishavebeenfastdevelopedbecauseoftheirversatilityandflexibility.Manywell-knownnumericalmethodshavebeenintroducedduringthattime,includingthefiniteelementmethod(FEM)[1]andthefinitedifferencetimedomainmethod(FDTD)[2,3],whicharebasedonthesolutiontotheMaxwell’sequationsindifferentialform,andthemethodofmoments(MoM)[2,4-6],whichisbasedonthesolutiontotheMaxwell’sequationsinintegralform.Especiallyfrom1990s,withthefastdevelopmentsofhighperformancecomputingsystems,thetheoriesandmethodsofcomputationalelectromagneticshavebeenadvanceddramatically.Theincreasesoftheclockspeedandthememorysizeofcomputersystemsandthedevelopmentsofhighlyefficientelectromagneticcomputingalgorithmsmakethenumericalmethodscapableofsolvingelectromagneticengineeringproblems.
…
1.3ContentsandInnovationsoftheThesis
BasedontheCalderόnrelationandtheCalderόnidentities,thisthesishasdevelopedseveralCalderόnpreconditioningtechniquesandinvestigatedtheirapplicationsintheintegral-equation-basedcomputationalelectromagneticmethods.TheresearchcontenthascoveredtheCalderόnpreconditioningtechniquesfortheperfectelectricconductor(PEC)anddielectriccases.ForthePECcase,theCalderόnpreconditionsatmid,low,andhighfrequenciesareinvestigated.Forthedielectriccase,theCalderόnpreconditioningtechniquesforthePMCHWTandN-Müllerintegralequationsaredeveloped.Thenumericalaccuracyofthesecond-kindFredholmintegralequationsareinvestigatedandimprovedinthisthesis.
…
1.4OutlineoftheThesis
Thisthesisisorganizedasfollows.
…
Chapter2TheoreticalBasics
Inthischapter,thegeneralmethodsofconstructingthecommonlyusedintegralequationsinelectromagneticsareintroducedbasedonthesurfaceequivalenceprincipleandthevolumeequivalenceprinciple.
…
2.1IntegralEquationsinElectromagnetics
Intheintegral-equation-basedcomputationalelectromagneticmethods,theunknownfunctionsintheelectromagneticproblemssuchasthescatteringorradiationfieldsaremodeledintermsoftheequivalentsurfaceorvolumeelectric/magneticsourcesbyapplyingthesurfaceorvolumeequivalenceprinciples,respectively.
…
2.2270MHzPlanWaveExcitation
Inordertoinvestigatetheitsperformanceinhandlingelectricallyverylargeproblemswithoveronemillionunknowns,thesamenumericalexampleisrepeatedagainbyincreasingthefrequencyto270MHz,andkeepingtheincidentangleandpolarizationoftheplanewaveunchanged.Tohaveabetterinsight,thememoryconsumptionandCPUtimerequirementsoftheEFIE,theCP-CFIE(0.8),andtheCP-AEFIEalgorithmsaregiveninTable2-1.
Table2-1ComparisonofComputationalDataofDifferentAlgorithms
TotalMemory(Mb)
CPUTime
Setup(h)
SolutionTime
Iter.(m)
Tol.(h)
EFIE
3215.84
1.14
3.18
>63
CP-CFIE(0.8)
6386.12
7.84
7.04
27.69
CP-AEFIE
5750.43
6.71
7.47
19.05
AllthecalculationsarecarriedoutonaHPZ400workstationwithaFedora10operatingsystem.
…
2.3TheSolutionofIntegralEquationsinElectromagnetics
2.3.1GeneralPrincipleoftheMethodofMoments
Theintegralequationsconstructedintheprecedingsectioncanbesolvedwithadequatenumericalmethods.Oneofthemostcommonlyusedmethodsinsolvingintegralequationsisthemethodofmoments(MoM)introducedbyR.F.Harringtonin1968[5].ThegeneralprincipleandkeypointsofMoMwillbereviewedinthissection.
…
2.3.2GeometricalModelingandDiscretizationofObject
Fromthedescriptionintheprecedingsection,itisclearthatinordertosolvefortheunknownequivalentelectromagneticcurrentsdefinedonthesurfaceorinthevolumeofanobstruction,thedefinitiondomainoftheunknowncurrents,whichisthegeometry,needstobedescribedmathematically.Thisistheso-calledgeometricalmodeling.Incomputationalelectromagnetics,geometricalmodelingisthebasicofelectromagneticmodelingandnumericalcalculation,anditsqualitywillaffecttheaccuracyofthenumericalsolutiondirectly.
…
2.3.2.1PlanarTriangularModel
Thesimplestandmostcommonlyusedelementinthegeometricalmodelingistheplanartriangle,whichisdefinedbyitsthreevertices(nodes).
…
2.3.2.2CurvilinearTriangularModel
Thecurvedsurfaceofanobjectcanbebettermodeledwithcurvilineartriangularelementswhicharethesecond-ordercurvedsurfaces.Acurvilineartrianglecanbedefinedbysixnodes,threeofwhicharetheverticesofthetriangle,theotherthreearethemidpointsofthreecurvededges.ShowninFigure2-1isthesketchofacurvilineartriangularelement.
Thecurvedsurfaceofanobjectcanbebettermodeledwithcurvilineartriangularelementswhicharethesecond-ordercurvedsurfaces.Acurvilineartrianglecanbedefinedbysixnodes,threeofwhicharetheverticesofthetriangle,theotherthreearethemidpointsofthreecurvededges.ShowninFigure2-1isthesketchofacurvilineartriangularelement.
Figure2-1Thesketchofacurvilineartriangularelement.(a)Thecurvilineartriangleinthecoordinatesystem;(b)Thecurvilineartriangleinthecoordinatesystem
Usingthefollowingcoordinatetransformation,thecurvilineartriangleintherectangularcoordinatesystem,asshowninFigure2-1(a),canbemappedontothetriangledefinedinaparametriccoordinatesystem,asshowninFigure2-1(b)
(2-1)
where
denotestherectangularcoordinatesofthesixcontrollingnodesinFigure2-1a,
,
,
aretheparametriccoordinatesvaryingfrom0to1,andtheysatisfytherelation
(2-2)
From(2-2),itisclearthatonlytwovar