简介：Byexpandingtheperturbationofcovariancematrixinthepowersofer-rorterm,theinfluencefunctionsforfivecanonicalmeasurementsinCCAaredevel-opedandthreesampleversionsaregiven.Forgeneralizedcorrelationcoefficientp_z,theinfluencefunctionisaquadraticformofr.v.z,anditsdistributionisconsidered.ApracticalexampleiUustratestheutilityoftheproposedinfluencefunctions.

简介：

简介：Inordertoobtainthesolubilitydataofpropylenein2-propanol-watersolutions,gas-liquidequilibrium(GLE)experimentswerecarriedoutat303.15-333.15K,0.27-0.98MPainastaticequilibriumstill.Theoriginalmassratioof2-propanoltowaterwas9∶1,8∶2,7∶3,1∶0,respectively.TheequilibriumdatawerecorrelatedwithanempiricalmodelandtheregressionofmodelparameterswascompletedbyGauss-Newtonnonlinearleastsquare(NLS).Theaveragerelativedeviation(ARD)betweentheexperimentalandcalculatedvalueis1.5700,andthemaximumrelativedeviation(MRD)is4.8200.Inaddition,asimpleapproachthatcorrelatedthemodelparameterswiththesystemcompositionwasalsoprovided.

简介：Thestressintensityfactor(SIF)isacriticalparameterassociatedwiththefracturebehaviourofmaterials.Inthispaper,weselectthedisplacementfunctionaroundacracktipastheshapefunctionofthedigitalimagecorrelation(DIC),whichmakesitpossibletodirectlycalculatetheSIFbythecorrelationscheme.Moreover,weuseanon-rectangularsubset,whichcanreducetheinfluenceofplasticdeformationandcrackwidthontheDICmeasurementaccuracy.WemeasuredtheSIFofamodeIcrackinasuper-hardaluminiumalloyspecimentoverifytheperformanceoftheproposedmethod.OurexperimentalresultsshowthataDICwithaspecificshapefunctioncanbeusedtoaccuratelyandefficientlycalculatetheSIF.Furthermore,wealsopresentapracticalapplicationofourproposedmethodfordeterminingtheSIF,crackpropagationangleandcracktipdisplacement.

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：天体的跨matching是为聚集不同波长的观察数据的一个基本方法。由数据聚集，天体的目标的性质能包括地被理解。在减少瞄准时间在I/O上消费了操作，几个改进方法被介绍，基于养模特儿的边界包括处理流动，它能减少数据库询问操作；能改进任务分区的表演并且解决数据稀少的问题的最大的成长块和它的决心的一个概念；并且计算这个索引的一个快bitwise算法附近的块数，它是一条重要效率保证。实验证明方法能有效地在稀少的数据集和高密度的数据集上加快跨matching。