结合初值选取的最大似然估计被动定位算法（英文）.pdfVIP

第4O卷第3期 光电工程 Vo1．4O．No．3 2013年 3月 Opto—ElectronicEngineering March，2013 文章编号：1003—5OLX(2013)03—0007—07 M aximum LikelihoodEstimationAlgorithm Combiningwith InitialValueSelectionforPassiveLocalization CHENJinguang1,2 M A Lili 1 (1．SchoolofCompu~rScience，XianPolytechnicUniversity，XJan710048，China； 2．SchoolofElectronicEngineering,XidianUniversity,Xian710071，China) Abstract：TheproblemsofconventionalMaximumLikelihoodEstimation(MLE)algorithm areaddressedforpassive localization andan improvedMLE passivelocalization algorithm ispresneted．Firstly,weestimatean initialtarget positionusingleastsquraemethod．Moreover,inordertoadaptthemeasurementerror,thesquraerootofthedifference betweentheestimatedpositionandthesensor’spositionisusedastheapproximatecovariancematrixofrmeasurement error．Then，aweightedleastsquareformulaisemployedtoestimateanew position．Finally,weregradhteestimation valueasthenecessary initialvalue，andemploytheconventionalMLEtocalculatethefinalresults．Theimproved algoritmh havesomeadvantages，i．e．，itdoesnotneedtoset811initialatrgetposition，itslocalizationresultsdOnotdiverge easily,itscompuattionalcomplexity islOW,nad ithasthe same levelin accuracy as htatofconventionalMLE． Experimentalresultsshowthattheproposedalgorithm iseffective． Keywords：passivelocalization；targetlocalization；mxaimmu likelihoodestimation；adatfusion CLCnumber：TP391 Documentcode：A doI：10．3969~．issn．1003-501X．2013．03．002 结合初值选取的最大似然估计被动定位算法 陈金广 ，马丽丽 (1．西安工程大学 计算机科学学院，西安 710048，