心臟信號混沌時間序列分析及相關軟件研究.pdf

1、同濟大學碩士學位論文心臟信號混沌時間序列分析及相關軟件研究姓名：易攀科申請學位級別：碩士專業(yè)：控制科學與工程指導教師：蔣式勤20050201同濟大學申請碩士學位論文摘要ABSTRACTHumanheartisaverycomplicatedphysiologicalsystemThemethodofdetectingandanalyzinghumanheartsignalaccordingtocomputergraphictheoryh

2、asbeentheproductioninthefieldofhumanlifescience，suchasMagnetoCardioGram(MCG)andMagneticResonanceImaging(MRI)Comparewiththetraditionalelectricsignalofonrheart，themagneticsignalisdetectedwithoutcontacting，beingmoreinsuscep

3、tibletothehumanphysicalorganizationandmoresensitivetotheelectriccurrentof‘static’heartmuscleThemagneticsignalisatypicalnonlineartimeserialbasedontheviewoftimeserialanalysistheoryInordertostudytheessenceofhumanheartandpro

4、videusefulproofforthediagnosisofheartdiseases，thispaperstudiesthenonlinearcharacteroftheheartmagneticsignalandtheheartelectricsignalbyuseofchaoticnonlineartimeseriesanalysismethodsFirstly，theattractorofthemagneticsignala

5、ndtheelectricsignalareconstructedrespectivelyTimedelayischosenbyusingmutualinformationmethod，whilereconstructdimensionisobtainedbyCaomethodbasedonfalsenearestneighbormethodSecondly，therespectivecorrespondingrelationshipb

6、etweenthethreeloopsoftheattractoroftheheartelectricsignalandthethreewavesinElectroCardioGram(ECG)，namely，Pwave、QRSwaveandTwave，isprovedbyimitationforthefirsttimeThirdly，thefeatureoftheattractorofthemagneticsignalandtheel

7、ectricsignaliSanalyzedFractaldimensioniSestimatedbyG—Psaturationcorrelationdimensionmethod，andmaximalLyapuonvexponentiscalculatedbyforecastingexponentmethodbasedonWolfmethodFourthly，theonerankaddingweightlocalregionmetho

8、dispropoundedandusedinthepredictionofthetimeseriesofthemagneticsignalandtheelectricsignalTheresultstestifythedifferenceofmaximalLyapuonvexponentsofthetwokindsofsignalFinally，severalimportantmodulesinMCGsoftwarearedevelop

9、edTheaboveworkshowsthattheheartmagneticsignalmadthehearteleetricsignalbothreflecttheelectricalactivityofOurheart，andithasimportanttheorysignificanceandgoodreferencevalueforfurtherstudiesofchaoticattractorsofdifferenthear