欢迎来到天天文库
浏览记录
ID:36837954
大小:3.47 MB
页数:70页
时间:2019-05-16
《用于机载平台机动目标跟踪的多传感器交互式多模型数据概率关联算法的研究》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、摘要在军用和民用领域,机动目标跟踪始终是人们的密切关注的重要问题之一。随着数据融合理论的不断发展和广泛应用,将数据融合用于机动目标跟踪已成了研究的必然发展趋势。在运用了数据融合的机动目标跟踪算法中,概率数据关联(PDAF)算法在杂波环境下有很好的跟踪性能,而多模型算法(MMF)适用于目标高度机动的情况,由此,导出了一种适用于杂波环境的对机动目标跟踪的交互式多模型——概率数据关联(mm卜—PDAF)算法。在此基础上,本文采用了脉冲多普勒(pD)雷达、电子支援系统(ESM)和敌我识别系统(IFF)三个机载异类传
2、感器,在选取适用的运动模型和坐标系之后,运用多传感器交互式多模型(IMM--MSPDAF)算法,对不同的机动和非机动目标分别进行了三维跟踪仿真,并与单传感器单模型跟踪算法的跟踪仿真进行了对比分析。结果表明:由于多传感器交互式多模型概率数据关联(IMM--MSPDAF)算法充分利用了各传感器的各自特性弥补了单传感器对目标探测的不足,运用多模型的并行交互作用解决了单一模型在跟踪机动目标时的缺陷,从而很好的实现了在密集杂波环境下对机动目标的跟踪,明显提高了跟踪系统的跟踪精度。关键词:机动目标跟踪交互式多模型数据融
3、合异类多传感器概率数据关联算法ABSTRACTManeuveringtargettrackingisoneofimportantproblemswhichreceivecloseattentioninthedefensefieldsandCivilfields.Wjththecontinuousdevelopmentandwideapplicationofdatafusiontheory,itisallinevitableresearchtrendofusingdatafusionintomaneuveri
4、ngtargettracking.Inthemaneuveringtargettrackingalgorithmusingdatafusion,probabflisticdamassociationalgofithmhasperfecttrackingabilityinclutterenvironment,andmultiple-modelalgofithmissuitableforthecaseoftrackingtargetwithhi曲maneuvering,therefore,probabilist
5、icdataassociationalgorithmisdeduced.whichistheinteractingmuldplemodelandsuitablefortrackingmaneuveringtargetinclutterenvifoment.Onthesebases,threedifferenttypesofairbornesensols,namelyPulsedDopplerRadar,ElectronicSupportMeasurementandIdentificationbetweenF
6、riendurFoeareadopted.After埔趣edsuitablemotionmodelsandcoordinates,multi-sensorinteractingmultiplemodelprobabilisticdataassociationalgorithmisappliedtotrackdifferentmaneuveringtargetsandnon-maneuveringtargetsinthreediInension&Atthesarnetime,theresultsoftrack
7、ingsimulationalecomparedandanalyzedwiththosetrackedbysinglesensorandsinglemodeltrackingalgorithm.Theresultsshowthatmulti-sensorinteractingmultiplemodelprobabilisticdataassociationalgorifluncantrackthemaneuveringtargetperfectlyinclutterenvironmentandimprove
8、trackingprecisionobviously,becausethealgorithmmakesfullUseoftheinformationofeachsensorfullytomakeuptheshortageoftargetdetectionofsinglesensorandUSestheparallelinteractionofmodelstosolvethedefectionoftrackingm
此文档下载收益归作者所有