您好,欢迎访问三七文档
当前位置:首页 > 商业/管理/HR > 信息化管理 > 信息融合理论的基本方法与进展-潘泉
2910201210ControlTheory&ApplicationsVol.29No.10Oct.2012():1000¡8152(2012)10¡1233¡12,,,,(,710072):,,,.,;,10.,.:;;;(JDL):TP273:ABasicmethodsandprogressofinformationfusion()PANQuan,WANGZeng-fu,LIANGYan,YANGFeng,LIUZhun-ga(SchoolofAutomation,NorthwesternPolytechnicalUniversity,Xi’anShaanxi710072,China)Abstract:Becauseofthedemandsfrommilitarytechnology,automationandintelligence,informationfusionhasbeenbroadlyconcernedinacademiaandindustry.Sinceanumberofmethodsandalgorithmshavebeenemerginginrecentyears,itisnecessarytoanalyzeandreviewtheirrecentdevelopments.Westartbyanalyzingtheinherentproblemsandchallengesininformationfusion,includingfusionarchitecture,uncertaintyofinformation,multimodalinformation,conflictinformation,correlativeinformation,networkedinformationandnonlinearityofinformation;andthen,wereviewvariousmethodsandalgorithmsdevelopedinthepasttenyearsforimplementingthefusionoftheabovetypesofinformation.Inaddition,severalfuturedirectionsofresearcharehighlightedanddescribed,includingthejointprocessingforinforma-tion,human-centeredinformationfusion,jointoptimizationforinformationgatheringandfusion,architecturedesignforcomplexmultisensorinformationfusionsystems,simulationandperformanceevaluationforinformationfusionandtheapplicationofmoremathematicaltheories.Keywords:informationfusion;datafusion;multisensorsystem;JointDirectorLaboratory(JDL)model1(Introduction),.,,[1].C3I(command,control,communicationsandintelligence),2070,,.,.,,.:,().,,,.,,.,:,;,;,;:2011¡11¡22;:2012¡08¡08.:(61135001,61074179,61075029);(cx201014).123429,.,,..[2].10,,,.[2],,10,;,.,,.,.2...2(Problemsandchallengesofinformationfusion),,.:1):,,,.,;2):,;,.;3):,,,,;4)[3]:,.Dempster,.,;5)[3]:,,,,.,.,;6):,.,,;7):,..,.3(Methodsandalgo-rithmsofinformationfusion),.2,..,,(4).,InformationFu-sion2005,26,:Agent.[1][4–6][7][8–12]..3.1(Informationfu-sionmodelsandsystemdesign),,:JDL(JointDirectorsofLaboratories)DasarathyBoyd,,.,JDL,.10:()1235JDL4:(1);(2),1;(3),2,;(4),,,..JDLLOCE()LENSCE(),[13].Steinberg[14],JDL,0,,.Hall[15]Blasch[16]JDL5,,/..,.JDL1.5,,.PotomacFusionJDL(fusionexploitationframework,FEF),[17].1[18]Fig.1JDLinformationfusionmodel[18],COTS()[19].3.2(Uncertaininformationfu-sion).,.:,.:,.,,.,,,..,.4[20]:.(EKF)EKFEKF,.,.,.,.(Gaussiansumapproximation),,,,,.,.UKF[21](GHF)[20](CDF)[22](DDF)[22–23](particlefilter,PF).,,...PF,,.PF(),,(MonteCarlo,MC).,PF.,,PF,,(MCMC)Unscented(APF)Rao-Blackwellised(RBPF)123629(RPF)[24].PF.,MCMC,.,.MCMC,Metropolis-Hastings[25]Gibbs[26].Metropolis-Hastings,,.GibbsMetropolis-Hastings.Rao-Blackwellised[27],PF(marginalizedparticlefilters),,,,,PF,,PF,[28–29].,[18–21,30–31].:Kendall[32]2070.Goodman[33].Mahler[34],.[35].,,.,,.,//,..(PHD),PHD.(CPHD).,,MC.(C)PHD,[36].PHD(),./.,.[37]:,,,,/.,.3.3(Multi-modalinformationfu-sion).,.,.,.,,,.,[38].,.,.,.:,,[39];,,;,.,.:,;;,,.,.,,.,[38].,.,[40].:Dempster-Shafer(DS).10:()1237,.EKFPF[41],.0–13–D..,.4.3.4(Conflictinformationfusion),.,Dempster.,[42–46][47–51].,,.,,,.[52–54],[47,55–56].,,,.[57],.,[58].DSmT(Dezert-Smarandachetheory)[42]DS,DS,,DS,.,,JeanDezertPCR1–PCR6[42]Shafer,.,.Uhlmann[59](covarianceunion),.Maskell[60]PFMCMC,DSmT,.,,.,,,[61–62].3.5(Correlativeinformationfusion),.,,;,,.,,.,.,;,.,,.,,,,,.,.(covarianceintersection,CI).,.,CI,.,CI.,,:;CI,;,;Kalman,PF.,Benaskeur[63],CI.[64][63],.CI,Farrell[65]Chernoff,.Bar-shalom[66]–,,,–.,..,.,.,(latentsemanticanalysis)(canonicalcorrelationanalysis)(cross-modalfactoranalysis)(causallinkanalysis)(causalstrength)(agreementcoefficient).,5,3.123829,..,,[38].,Sundaresan[67]Copula,,,().3.6(Networkedinformationfu-sion),(networkeddatafusion).,.,,[68].,:()()()[69–72].,,.,.Fletcher[73],,,.Sinopoli[74],.Yue[75]().Gao[76],.Liang[77]––,,,,.,.:;,,,,.,[78].Benveniste,Willsky,Hong,2,;,,.,Liang[79],[80].,,,–.Shen[81–82],..,,,,(out-of-sequencemeasurements,OOSM).OOSM,.,,,.,OOSM.OOSM,OOSMOOSMOOSM.OOSMBar-ShalomABC[83],Zhou[84]A.Zhang[85]A,OOSM.Bar-Shalom[86],ABAl1Bl1,Bl1,.Oreshki[87]PF,,OOSM.Shen[88–90],,;,10:()1239,.OOSM,,,,OOSM,.Mallick[91]OOSM,OOSM.Maskell[92],OOSMOOSM.OOSM,.,OOSMOOSM.4(Emerginganddevelopingparadigms),.,,..4.1(Jointprocessingforin-formation),,,,,Neyman-Pearson,.,..,,,,..,.,[93–96][97–102].4.2(Humancenteredin-formationfusion),,,,,.,().,,..,,.,.,[18,103–107].,,;,,;,,.,.[108].4.3(Jointoptimiza-tionforinformationgatheringandfusion),,....,(),.[109–112],,,.,,:(controlofinformationcollectionandfusion,2010)(valueofinformationfordistributeddatafusion,2011).4.4(Architecturedesignforcomplexmulti-sensorinformationfusionsystems).–.,,(),,124029,.,,..DoDAF(departmentofdefensearchitectureframework)C4ISR(command,control,com-munications,computers,intelligence,surveillanceandreconnaissance),C4ISR,,C4ISR[113],.,DoDAF,,.UML,sysML,IDEF,,,[114–116].,.DoDAF,.4.5(Simulationandperformanceevaluationforinformationfu-sionsystems),,,,.:,;,;,.,.,..4.6(Leveragingmoremathematicaltheories),,.,.,,,.,,[117].,.–[118],[119–120].Sensor-Rich,.(References):[1],,.[M].2.:,2010.(HANChongzhao,ZHUHongyan,DUANZhansheng.Multi-SourceInformationFusion[M].2ndEdition.Beijing:TsinghuaUniversityPress,2010.)[2],,,
本文标题:信息融合理论的基本方法与进展-潘泉
链接地址:https://www.777doc.com/doc-1746220 .html