简介：本文首次给出了非线性随机效应模型参数估计弱相合性的必要条件，并证明了在一定正则条件下估计的强相合性和渐近正态性，推广了ＢｒｅｓｌｏｗａｎｄＣａｌｙｔｏｎ（１９９３），ＬｅａｎｄＮｅｌｄｅｒ（１９９６）的结果．

简介：Fromthegroupmovementofthebedloadwithinthebottomlayer,detailsofthenonlineardynamiccharacteristicsofbedloadmovementarediscussedinthispaper.Whetherthesedimentisinitiatedintomotioncorrespondstowhethertheconstanttermintheequationisequaltozero.Ifconstanttermiszeroandnodispersiveforceisconsidered,theequationrepresentsthetraditionalShieldsinitiationcurve,andifconstanttermiszerowithoutthedispersiveforcebeingconsidered,thenanewShieldscurvewhichismuchlowerthanthetraditionaloneisgot.Thefixedpointoftheequationcorrespondstotheequilibriumsedimenttransportofbedload.Inthemutationanalysis,wehavefoundthattheinflectionpointisthedemarcationpointofbreaking.Intheory,thebreakingpointcorrespondstothedividingboundaryline,acrosswhichthebedformchangesfromflatbedtosandrippleorsanddune.ComparedwiththeexperimentaldataofChatouHydraulicLabinFrance,theconclusionsareverified.

简介：Basedonphase-planeorbitanalysis,themathematicalmodelofpiecewise-smoothsystemsofmulti-degree-of-freedomunderthemodecoordinateisestablished.Approximateanalyticalsolutionunderthephysicalcoordinateofmulti-degree-of-freedomself-excitedvibrationinducedbydryfrictionofpiecewise-smoothnonlinearsystemsisderivedbymeansofaveragemethod,theresultsofwhichagreewiththoseofthenumericalsolution.Aneffectiveandreliableanalyticalmethodinvestigatingpiecewise-smoothnonlinearsystemsofmulti-degree-of-freedomhasbeengiven.Furthermore,thispaperqualitativelyanalysesthecurvesaboutstationaryamplitudeversusrubbingvelocityofhandsandversusnaturalfrequencyofhands,andaboutangularfrequencyversusrubbingvelocityofhands.Theresultsprovideatheoreticalbasisforidentifyingparametersofthesystemandtheanalysisofsteadyregion.

简介：为为非线性的系统的适应基于观察员的差错估计的一条途径被建议。H无穷理论被用于分析设计方法和适应观察员，系统状态和差错能从被估计的稳定的条件。差错估计错误与给定的H无穷磁道性能索引有关，这被证明，以及差错和非线性的条款的Lipschitz常数评价到变化。适应观察员的设计步被建议。甚至当系统包括非线性的条款时，模拟结果证明观察员为差错估计有好性能，它证实方法的有效性。

简介：滑动模式控制的适应backstepping与输入浸透为不明确的非线性的系统的一个班被建议。一个命令过滤了途径被用来阻止输入浸透破坏神经网络(NN)的适应能力。控制法律和NN的适应更新法律在Lyapunov功能的意义被导出，因此稳定性能甚至在输入浸透下面被保证。建议控制法律对骚乱是柔韧的，并且它能也消除输入浸透的影响。模拟结果显示建议控制器有好表演。

简介：Thispaperdiscussesthestabilityoftheoreticalsolutionsfornonlinearmulti-variabledelayperturbationproblems(MVDPP)oftheformx′(t)=f(x(t),x(t-τ1(t)),…,x(t-τm(t)),y(t),y(t-τ1(t)),…,y(t-τm(t))),andgy′(t)=g(x(t),x(t-τ1(t)),…,x(t-τm(t)),y(t),y(t-τ1(t)),…,y(t-τm(t))),where0<ε<<1.Asufficientconditionofstabilityforthesystemsisobtained.AdditionallyweprovethenumericalsolutionsoftheimplicitEulermethodarestableunderthiscondition.

简介：Withthetermsoftheexactseriessolutiontakenastrialfunctions,themethodofpointcollocationwasusedtocalculatethelargedeflectionofacircularplate.Theaxisymmetricalbendingformulaeweredevelopedforthecalculationofacircularplatesubjectedtopolynomialdistributedloads,aconcentratedloadatthecenter,uniformradialforcesandmomentsalongtheedgeortheircombinations.Thesupportmaybeelastic.Thebucklingloadwascalculated.Underactionofuniformlydistributedload,centralloadortheircompoundload,solutionswerecomparedwiththoseobtainedbyothermethods.Bucklingbeyondcriticalthrustwascomparedwiththatcalculatedbythepowerseriesmethod.Themethodpresentedinthispaperhasadvantagesofwideconvergentrange,highprecisionandshortcomputingtime.Moreover,thecomputingtimeisnearlyindependentofthecomplexityoftheloads.

简介：一个基于核的判别式分析方法把核称为直接判别式分析被采用，它把直接线性的判别式分析的优点与核诡计的相结合。为了表明它的更好的坚韧性到真实的脸的复杂、非线性的变化，想象，例如照明，面部表情，规模和姿势变化，实验在Olivetti研究实验室，耶鲁和自我造的脸上被执行数据库。Theresults与核主管部件分析和核相对照显示那线性判别式分析，方法能完成更低(7%)用仅仅特征的一个很小的集合的错误率。而且，一个新改正的内核模型被建议改进识别性能。试验性的结果证实它的优势(1%以识别评价)到另外的多项式，核当模特儿。

简介：本文提出一种交互参考方法，用于半盲情况下非线性时间序列的分析，这时描述该序列的动力学方程形式已知而相应的参数未知．通常是分别完成的噪声抑制和参数估计这２个任务在这里被组合在一起迭代完成．由于２个处理模块间的积极相互作用，该方法可获得更好的性能．以前的一些分析方法可以看作是这一处理框架的特例．将其用于含噪声混沌时间序列的分析，实验结果表明了该方法对性能的显著改善．

简介：Anewmethodforidentifyingnonlineartime-varyingsystemswithunknownstructureispresented.Themethodextendstheapplicationareaofbasissequenceidentification.Theessentialideaistoutilizethelearningandnonlinearapproximatingabilityofneuralnetworkstomodelthenon-linearityofthesystem,characterizetime-varyingdynamicsofthesystembythetime-varyingparametricvectorofthenetwork,thentheparametricvectorofthenetworkisapproximatedbyaweightedsumofknownbasissequences.Becauseofblack-boxmodelingabilityofneuralnetworks,thepresentedmethodcanidentifynonlineartime-varyingsystemswithunknownstructure.Inordertoimprovethereal-timecapabilityofthealgorithm,theneuralnetworkistrainedbyasimplefastlearningalgorithmbasedonlocalleastsquarespresentedbytheauthors.Theeffectivenessandtheperformanceofthemethodaredemonstratedbysomesimulationresults.

简介：AmodifiedLindstedt-Poincaré(LP)methodforobtainingtheresonanceperiodicsolutionsofnonlinearnon-autonomousvibrationsystemsisproposedinthispaper.Inthemodifiedmethod,nonlinearnon-autonomousequationsareconvertedintoagroupoflinearordinarydifferentialequationsbyintroducingasetofsimpletransformations.Anapproximateresonancesolutionfortheoriginalequationcanthenbeobtained.Theperiodicsolutionsofprimary,super-harmonic,sub-harmonic,zero-frequencyandcombinationresonancescanbesolvedeffectivelyusingthemodifiedmethod.Someexamples,suchasdampedcubicnonlinearsystemsundersingleanddoublefrequencyexcitation,anddampedquadraticnonlinearsystemsunderdoublefrequencyexcitation,aregiventoillustrateitsconvenienceandeffectiveness.UsingthemodifiedLPmethod,thefirst-orderapproximatesolutionsforeachequationareobtained.Bycomparison,themodifiedmethodproposedinthispaperproducesthesameresultsasthemethodofmultiplescales.