帶乘性噪聲系統(tǒng)最優(yōu)估計(jì)的數(shù)值穩(wěn)定性方法研究.pdf

1、中國(guó)海洋大學(xué)碩士學(xué)位論文帶乘性噪聲系統(tǒng)最優(yōu)估計(jì)的數(shù)值穩(wěn)定性方法研究姓名：陳萌申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：信號(hào)與信息處理指導(dǎo)教師：褚東升20040601StudiesofNumericaIIyStabIeEstimationforMuIti—OhanneISystemswithMultiP1icativeNoiSeSABSTRACTOptimalestimationtheoryformulti—channelsystemswithmultip

2、licativenoisesisveryimportantinmanyapplicationssuchasoilseismicexploration，underWaterremotetargetsdetectionandspeechsignalprocessingInrecentyearstherehadbeengrowingresearchinterestsandgl’vingmanynewalgorithmsinthisfieldT

3、hepreviousresearchesintheestimationtheoryarebasedonthesystemswithmultiplicativenoises(SMN)，recentyearspeoplepaidmoreattentiontothemulti—channelsystems(MSMN)Manyoptimalestimationalgorithmsaregiven，buttheserecursivealgorit

4、hmsarenumericallyunstable，itmeansthatthcywillbecomeinvalidsometimeswhenuseittosolveproblemsSo，newmethodswhichhaveexcellentnumericalstabilityarenecessaryAlltheproposedalgorithmsinthisdissertationareoptimalinthesenseofline

5、arminimum—vailaliceNumericallystablemethodsforMSMNareproposedinthispaperThemaincontentsofthisdissertationareasfollows：1ThedevelopmentandstatusquoofsignalestimationforMSMNisrecalledMoreover，numericalstabilitytheoryisintro

6、ducedsimply2GiveanSVDbasedoptimalfilteringforMSMN，ithasanexcellentnumericalstabilityItisoptimalstillintheSCD_SCoflinearminimumvariance3OptimaismoothingalgorithmsforMSMNareproposed，thennumericallystablealgorithmsbasedonsi

7、ngularvaluedecomposition(SVD)methodsaregivenforsmoothingDeconvolutionalgorithmsareproposedsubsequentlybasedontheresultsoffilteringandsmoothingalgorithmsncyarestilloptimalinthesenseoflinearminimum—variance4Simulationexamp