EM算法在雜波環(huán)境下機(jī)動目標(biāo)跟蹤中的應(yīng)用研究
Study of Application EM Algorithm on Tracking Maneuvering Targets with Clutter
-
摘要: EM(Expectation-Maximization)作為一種迭代求解非完備數(shù)據(jù)條件下極大似然(后驗(yàn))參數(shù)估計(jì)問題的方法,在目標(biāo)跟蹤領(lǐng)域主要應(yīng)用于被動跟蹤及實(shí)時性要求不高的目標(biāo)環(huán)境.該文推廣了L.A.Johnston的理論成果,推導(dǎo)得出了一種基于AECM(Alternative Expectation ConditionMaximization)方法的雜波環(huán)境下實(shí)時機(jī)動目標(biāo)跟蹤箅法,算法中后驗(yàn)?zāi)Ｐ透怕逝c關(guān)聯(lián)概率由隱馬爾科夫模型濾波計(jì)算得到.仿真計(jì)算表明,所提算法跟蹤精度與IMM-PDA性能相當(dāng),算法是有效的.
-
關(guān)鍵詞:
- EM算法;隱馬爾可夫模型;交互式多模型箅法;機(jī)動目標(biāo)跟蹤
Abstract: The EM algorithm, as an iterative numerical tool for computing maximum likelihood (or MAP) parameter estimates for incomplete data problem, has been used in area of target tracking, particularly in passive tracking and scenario in which real-time processing is unnecessary. As an extension of .Johnstons recent work, a recursive algorithm for tracking maneuvering targets in clutter, which based on AECM algorithm, is developed in this paper. In this algoritlim, model posterior probability and data association probability are computed via HMM filter respectively. Computer simulation indicates that performance of the algorithm is comparable with that of IMM-PDA, and the algorithm is valid. -
Bar-Shalom Y,Chang K C,Blom H A P.Automatic Track Formation in Clutter with A Recursive Algorithm.MA,USA,Artech House INC,1990:25-42.[2]潘泉,劉鋼,戴冠中,張洪才.聯(lián)合交互式多模型概率數(shù)據(jù)關(guān)聯(lián)算法.航空學(xué)報,1999,20(3):234-238.[3]Blackman S S,Bush M T,Populi R F.IMM/MHT tracking and data association for benchmark problem.Proc.of American Control Conference,Seattle,WA,1995:2606-2610.[4]Lang Hong,Ningzhou Cui,Shan Cong.Devert Wicker,An interacting multipattern data association (IMPDA) tracking algorithm[J].Signal Processing.1998,71(4):55-77[5]Logothetis A,et al..A Bayesian EM algorithm for optimal tracking of a maneuvering target in clutter[J].Signal Processing.2002,82(3):473-490[6]Avitzour D.A maximum likelihood approach to data association.IEEE Trans.on AES,1992,AES-28(2):560-566.[7]Dempster A,et al..Maximum likelihood from incomplete data viaEM algorithm.J.Roy.Stat.Soc.,1977,Sec39(12):1-38.[8]Johnston L A,et al..An improvement to the interacting multiple model (IMM) algorithm.IEEE Trans.on Signal Processing,2002,SP-49(12):2909-2923. -
計(jì)量
- 文章訪問數(shù): 2622
- HTML全文瀏覽量: 123
- PDF下載量: 614
- 被引次數(shù): 0
下載:
