简介:Thispaperisconcernedwiththeoptimallinearestimationproblemforlineardiscrete-timestochasticsystemswithrandommeasurementdelays.Anewmodelthatdescribestherandomdelaysisconstructedwherepossiblethelargestdelayisbounded.Basedonthisnewmodel,theoptimallinearestimatorsincludingfilter,predictorandsmootheraredevelopedviaaninnovationanalysisapproach.TheestimatorsarerecursivelycomputedintermsofthesolutionsofaRiccatidifferenceequationandaLyapunovdif...
简介:无线网络在更宽的光谱利用的时尚下面被开发(例如,认知无线电)并且多跳跃通讯(例如,无线网孔网络)。在这些范例,怎么有效地与最小化的相互的干扰分配光谱到不同传播连接成为关键担心。在这份报纸,我们在认知收音机网络(CRN)经由光谱分配学习产量优化。以前的研究合并冲突图或SINR模型描绘干扰关系。然而,以前的模型忽视积聚的干扰效果并且导致讨厌的干扰和非最优的结果,当工作在所有潜在的连接之中在估计的RSS(收到的信号力量)的精确性上基于后者模型忽视它的重信赖时。两个是不适当的描绘在干扰和产量之间的复杂关系。到这个目的,由考虑CR的特征,象光谱差异和间断OFDM一样,我们建议一个帮助测量的基于SINR的跨层的产量优化答案。我们的工作在不同的层使特征担心:在物理层,我们在场改进SINR模型的精确性的一个有效RSS评价算法;在上面的层,流动水平为WMN的基于SINR的产量优化问题作为一个混合整数被建模被证明NP难的非线性的编程问题。解决这个问题,一集中(1)最佳的算法和一个有效分布式的算法被提供。评估算法表演,真实世界的踪迹被用来说明我们的计划的有效性。
简介:Formultisensorsystems,whenthemodelparametersandthenoisevariancesareunknown,theconsistentfusedestimatorsofthemodelparametersandnoisevariancesareobtained,basedonthesystemidentificationalgorithm,correlationmethodandleastsquaresfusioncriterion.SubstitutingtheseconsistentestimatorsintotheoptimalweightedmeasurementfusionKalmanfilter,aself-tuningweightedmeasurementfusionKalmanfilterispresented.Usingthedynamicerrorsystemanalysis(DESA)method,theconvergenceoftheself-tuningweightedmeasurementfusionKalmanfilterisproved,i.e.,theself-tuningKalmanfilterconvergestothecorrespondingoptimalKalmanfilterinarealization.Therefore,theself-tuningweightedmeasurementfusionKalmanfilterhasasymptoticglobaloptimality.Onesimulationexamplefora4-sensortargettrackingsystemverifiesitseffectiveness.
简介:Magneticsensorarraysareproposedtomeasureelectriccurrentinanon-contactway.Inordertoachievehigheraccuracy,signalprocessingtechniquesformagneticsensorarraysareutilized.Simulationtechniquesarenecessarytostudythefactorsinfluencingtheaccuracyofcurrentmeasurement.Thispaperpresentsasimulationmethodtoestimatetheimpactofsensingareaandpositionofsensorsontheaccuracyofcurrentmeasurement.Severalerrormodelsarebuiltuptosupportcomputer-aideddesignofmagneticsensorarrays.
简介:Atemperaturecontrolsystemof31mverticalforcedair-circulationquenchfurnaceisproposed,whichisakindofequipmentcriticalforthermaltreatmentofaluminumalloycomponentsthatarewidelyusedinaerospaceindustry.Fortheeffectiveoperationofthefurnace,itisessentialtoanalyzetheradialtemperaturedistributionofthefurnace.Asetofthermodynamicbalanceequationsmodelingisestablishedfirstly.Byutilizingthenumericalanalysisresulttomodifythetemperaturemeasurements,thecontrolaccuracyandprecisionofthetemperaturearetrulyguaranteed.Furthermore,themultivariabledecouplingself-learningPIDcontrolalgorithmbasedonthecharacteristicsofstrongcouplingbetweenthemulti-zonesinthelarge-scaledfurnaceisimplementedtoensurethetruehomogeneityoftheaxialtemperaturedistribution.Finally,theredundantstructurecomposedofindustrialcontrolcomputersandtouchpanelsleadstogreatimprovementofsystemreliability.
简介:Anew3Dsurfacecontouringandrangingsystembasedondigitalfringeprojectionandphaseshiftingtechniqueispresented.Usingthephase-shifttechnique,pointscloudwithhighspatialresolutionandlimitedaccuracycanbegenerated.Stereo-pairimagesobtainedfromtwocamerascanbeusedtocompute3Dworldcoordinatesofapointusingtraditionalactivetriangulationapproach,yetthecameracalibrationiscrucial.Neuralnetworkisawell-knownapproachtoapproximateanonlinearsystemwithoutanexplicitphysicalmodel,inthisworkitisusedtotrainthestereovisionapplicationsystemtocalculating3Dworldcoordinatessuchthatthecameracalibrationcanbebypassed.Thetrainingsetforneuralnetworkconsistsofavarietyofstereo-pairimagesandthecorresponding3Dworldcoordinates.Thepictureelementscorrespondenceproblemissolvedbyusingprojectedcolor-codedfringeswithdifferentorientations.Colorimbalanceiscompletelyeliminatedbythenewcolor-codedmethod.Oncethehighaccuracycorrespondenceof2Dimageswith3Dpointsisacquired,highprecision3Dpointscloudcanberecognizedbythewelltrainednet.Theobviousadvantageofthisapproachisthathighspatialresolutioncanbeobtainedbythephase-shiftingtechniqueandhighaccuracy3Dobjectpointcoordinatesareachievedbythewelltrainednetwhichisindependentofthecameramodelworksforanytypeofcamera.Someexperimentsverifiedtheperformanceofthemethod.