简介:针对性能退化服从对称Laplace过程的产品,对其可靠性评估提出了贝叶斯方法.采用对称Laplace过程描述产品的退化过程,通过确定模型中的参数服从固定的先验分布,并且利用Gibbs抽样的方法,建立基于退化数据的贝叶斯估计模型,得到参数的后验分布.用蒙特卡洛模拟计算得到其可靠性,最后通过仿真模拟验证了模型的有效性.
简介:外汇收益率实际分布且有尖峰厚尾偏倚的特性,传统的正态分布不足以描述这一性质,而本文提出的非对称Laplace分布能够很好地拟合这一性质。非对称Laplace分布只有2个待估参数且存在有限的n阶矩,计算方便简单。深入分析了非对称Laplace分布的性质,对三种外汇数据进行了拟合,并做了拟合优度的K-S检验。结果表明,非对称拉普拉斯分能够比正态分布更好地描述外汇收益率的尖峰厚尾偏倚的特性。
简介:Inthispaper,thetheoremsconcerningthesummationofFourierserieswithparameteraregivenbyusingtheLaplacetransforms.BymeansoftheknownresultofLaplacetransforms,manynew,importantproblemsofsummationofFourierserieswithparameterinmechanicscanbesolved.
简介:AnewmethodforapproximatingtheinerseLaplacetransformispresented.WefirstchangeourLaplacetransformequationintoaconvolutiontypeintegralequation,whereTikhonovregularizationtechniquesandtheFouriertransformationareeasilyapplied.WefinallyobtainaregularizedapproximationtotheinverseLaplacetransformasfinitesum
简介:Inthepresentpaper,wehaveconsideredtheapproximationofanalyticfunctionsrepresentedbyLaplace-Stieltjestransformationsusingsequenceofdefiniteintegrals.WehavecharacterizedtheirorderandtypeintermsoftherateofdecreaseofEn(F,b)whereEn(F,b)istheerrorinapproximatingofthefunctionF(s)bydefiniteintegralpolynomialsinthehalfplaneRes≤b〈a.
简介:§0.IntroductionLetXbearealseparableBanachspaceandX*beitsdualspace,LetB(X)betheBorelfield,i.e.,thetopologicalσ-field.Afunctionalu:X→R’iscalledaboundedsmoothfunctional,ifn∈N,f1,…,fn∈X*andφ∈Cb∞(Rn),suchthat
简介:Letfibeaboundedsmoothdomain.Inthispaper,theauthorsdefinetheBesovspacesBpa,pon,establishtheatomicdecompositionofthesespaces,andobtaintheregularityestimateoftheDirichletproblemandtheNeumannproblemfortheLaplaceoperatoronthesespaces.
简介:Inthispaper,weconsidertheCauchyproblemfortheLaplaceequation,whichisseverelyill-posedinthesensethatthesolutiondoesnotdependcontinuouslyonthedata.AmodifiedTikhonovregularizationmethodisproposedtosolvethisproblem.Anerrorestimatefortheaprioriparameterchoicebetweentheexactsolutionanditsregularizedapproximationisobtained.Moreover,anaposterioriparameterchoiceruleisproposedandastableerrorestimateisalsoobtained.Numericalexamplesillustratethevalidityandeffectivenessofthismethod.
简介:Thispaperdevelopsanumericalmethodtoinvertmulti-dimensionalLaplacetransforms.Byavariabletransform,Laplacetransformsareconvertedtomulti-dimensionalHansdorffmomentproblemssothatthenumericalsolutioncanbeachieved.Stabilityestimationisalsoobtained.Numericalsimulationsshowtheefficiencyandpracticalityofthemethod.
简介:Fora〈r〈b,theapproachofLiandZhou(2014)isadoptedtofindjointLaplacetransformsofoccupationtimesoverintervals(a,r)and(r,b)foratimehomogeneousdiffusionprocessbeforeitfirstexitsfromeitheraorb.Theresultsareexpressedintermsofsolutionstothedifferentialequationsassociatedwiththediffusionsgenerator.Applyingtheseresults,weobtainmoreexplicitex-pressionsonthejointLaplacetransformsofoccupationtimesforBrownianmotionwithdrift,BrownianmotionwithalternatingdriftandskewBrownianmotion,respectively.