简介:新的戴尔PowerVault136T磁带存储系统基于新兴的线性磁带开放(LTO)标准,定位于那些有着大量数据备份需求的商业客户。PowerVault136T起价为人民币130,156元,拥有内置的系统管理诊断功能,支持高达14.4TB的存储容量、72个LTO磁鼓以及6个LTO磁带驱动器。为提供更快的速度和更多的用途,
简介:它的图表类型很多,科学图表和极坐标图表需要給PEP_faXDATA、PEP_faYDATA和PEP_faZDATA赋值,有以下的几个DLL函数可以为图表赋值
简介:在高速链路中,差分线由于具有很高的鲁棒性从而应用广泛,因此差分线的补偿就显得至关重要。以寻找最优的补偿方法为研究目的,结合实际PCB工作中的层叠结构,通过在HFSS和ADS中搭建模型,对比验证不同补偿方式对差分信号质量的影响。最后,指出存在的问题并总结归纳解决方法,为今后PCB布线中差分信号线的补偿方式提出了一些建议。
简介:Stationkeepingmode(SKM)forAttitudeOrbitControlSubsystem(AOCS);itisanoperationalapproachregardingtheattitudedynamicscontrol.Theessentialfeaturesofthedesignmethodologyisresearchingthebasictheoryandthenafterwards,makeaniterativedesignapproachusingsomepremise/assumptions,implementingthesystemsimulationthroughtheSimulinkMatlabsoftwarepackage,designingtherequiredcontrollers,monitorandanalyzingtheresponsesuntilthedesigngivethebestresultswithintherangerequired.Firstlythethrusterconfigurationisdesignedtoacquiretheparametersofthetorqueonthesatellite.Next,thecontrollersarebasedonthewellknownPIDcontrollaw.ControllersoptimizetheattitudeandareusedduringthemaneuversforSKM.Afterwardsthesimulationresultsaregiven,andpresentedtodemonstratetheperformanceandvalidityoftheAOCSdesignapproach.FinallythesimulationresultsdemonstratethatalltherequirementswereaccomplishedandtheStationKeepingModewassuccessfullydesigned.
简介:Althoughk-anonymityisagoodwayofpublishingmicrodataforresearchpurposes,itcannotresistseveralcommonattacks,suchasattributedisclosureandthesimilarityattack.Toresisttheseattacks,manyrefinementsofk-anonymityhavebeenproposedwitht-closenessbeingoneofthestrictestprivacymodels.Whilemostexistingt-closenessmodelsaddressthecaseinwhichtheoriginaldatahaveonlyonesinglesensitiveattribute,datawithmultiplesensitiveattributesaremorecommoninpractice.Inthispaper,wecoverthisgapwithtwoproposedalgorithmsformultiplesensitiveattributesandmakethepublisheddatasatisfyt-closeness.Basedontheobservationthatthevaluesofthesensitiveattributesinanyequivalenceclassmustbeasspreadaspossibleovertheentiredatatomakethepublisheddatasatisfyt-closeness,bothofthealgorithmsusedifferentmethodstopartitionrecordsintogroupsintermsofsensitiveattributes.Oneusesaclusteringmethod,whiletheotherleveragestheprincipalcomponentanalysis.Then,accordingtothesimilarityofquasi-identifierattributes,recordsareselectedfromdifferentgroupstoconstructanequivalenceclass,whichwillreducethelossofinformationasmuchaspossibleduringanonymization.Ourproposedalgorithmsareevaluatedusingarealdataset.Theresultsshowthattheaveragespeedofthefirstproposedalgorithmisslowerthanthatofthesecondproposedalgorithmbuttheformercanpreservemoreoriginalinformation.Inaddition,comparedwithrelatedapproaches,bothproposedalgorithmscanachievestrongerprotectionofprivacyandreduceless.