基于神經(jīng)網(wǎng)絡(luò)的多目標跟蹤信息融合技術(shù)的研究.pdf

1、哈爾濱理工大學碩士學位論文基于神經(jīng)網(wǎng)絡(luò)的多目標跟蹤信息融合技術(shù)的研究姓名：鄒海英申請學位級別：碩士專業(yè)：通信與信息系統(tǒng)指導(dǎo)教師：吳麗華20080301哈爾濱理工大學工學碩士學位論文ResearchonMultitargetTrackingInformationFusionBasedonNeuralNetworksAbstractDatafusionisarisingmultidisciplinarytechnologyItsapplic

2、ationtechnologydealswithawiderangeofdisciplinesandtechnologies，SOitisoneofthemostpotentialandvaluableresearchdirectionsintheworldtodayforitcanstrengthensystemviabilityreducetheinformationfuzziness，promotesystemreliabilit

3、yandSOonTargetTracking，whichisamajorpartofdatafusion，includessingletargettrackingandmulti—targettrackingInpractice，ittendstohavemanyindefiniteproblems，SOthatdataaresocomplexthatdataassociationandstateestimationbecomevery

4、difficultTherefore，it’SinevitabletointroduceNeuralNetworksandfuzzinesssystemintoMultitargetTrackingtechnologytomeetpeople’Sincreasingrequirementonbehaviorofdatafusionsystemsinceit’Shardtoworkjustdependingononlyonefusionm

5、ethodInthisdigsertation，multitargettrackinginformationfusionalgorithmbasedonNeuralNetworkshasbeenstudiedforthepurposeofpromotingsystemreliabilityandimprovingitsrobustnessFirstofall，thebasictheorems，topologyandalgorithmsi

6、ncommonuseofmultisensorinformationfusionhavebeenstudied，andtargettrackingtheoryanditsprocesshavebeendetailedintroducedFurthermore，Neuralnetworksbasedtrackingtheoryiscombinedwithfuzzinesstheoryinthispaper，andneuralnetwork

7、sapplicationsonmulti—targethavebeenresearcheddeeplySincethestudyrateofElmanneuralnetworkhasagreateffectuponconvergencespeedandstabilityanimprovedadaptivestudyalgorithmofElmannetworkisputforward，whichcanchangestudyrateacc