多通道帶乘性噪聲系統(tǒng)的觀測噪聲估計及狀態(tài)融合估計算法研究.pdf

1、中國海洋大學碩士學位論文多通道帶乘性噪聲系統(tǒng)的觀測噪聲估計及狀態(tài)融合估計算法研究姓名：王璐申請學位級別：碩士專業(yè)：信號與信息處理指導教師：褚東升20040601StudyofMeasurementNoiseandStateFusionEstimationAlgorithmsforMulti—ChannelSystemwithMultiplicativeNoisesABSTRACTOneoftheimportantprobleminsto

2、chasticsignalprocessingisoptimalestimationofsignalundercertainmeanings，orobtainingusefulinformationfromsignalspollutedbynoisesTheoptimalestimationalgorithmofmeasurementnoiseandtheoptimalstatefusionalgorithmformulticharme

3、lsvstemwithmultiplicativenoisesaremainlyresearchedinthisdissertationThisworkissupposedbyNSFSforTianyuanMathematicalFundGrant#A0324676andtheMinistryofEducationofChinaGrant#02131Atpresentthestudyofmulti—channelsystemwithmu

4、ltiplicativenoisesisnotperfectIn血efieldofnoiseestimationpreviousstudysareonly1imitedatdynamicnoiseestimation，butthemeasurementnoiseestimationhasalsoimportanttheoryvalueandpracticalsignificanceThisdissertationfocusesonthe

5、optimalestimationofmeasurementnoiseunderthecondifionthatthemultiplicativenoiseisintheforlTlofageneralstochasticmatrix，whichmeansagenuinemulti—charmelsystemFurthermomwjmtherapiddevelopmentofcomputertechnologytherehasbeena

6、considerableinterestinthemulti—sensorsystembecauseofitshighaccuracyandhighreliabilityAccordinglythemulti—sensorinformationfusiontechniquehasreceiveddeepresearchesandhasbeenusedinmilitarysystems，oilseismicexplorationandun

7、derwaterremotetargetsdetection，etcBasedontheformerworks，tllisdissertationalsodiscussestheoptimalfusionestimationalgorithmsThemainresearchingcontentsoftllisdissertationareasfollows：1Inthisdissertationanoptimalfilteringalg

8、orithmandanoptimalsmoothingalgorithmofmeasurementnoisearedevelopedformulti—channelsystemwithmultiplicativenoisesundertheconditionsthattheadditivenoisesareindependentwhitenoisesandthedynamicnoisecorrelatesitselfinone—step

9、andcorrelateswiththemeasurementnoiseatthepresentstepaswellasonepastste廿Thesealgorithmsareoptimaiinthesenseoflinearminimumvariance2Forthemultichannelsystemwithmultiplicafivenoisesundermulti—sensorobservationanoptimalfusio

10、nestimationaigorithrnisproposedontheconditionsthatthemultiplicativenoiseisintheformofageneralstochasticmatrixandthemeasurementnoisesofeachsensorarecorrelatedwiththedynamicnoiseThisexpandstheapplicationofmulti—sensorsyste