兩類隨機(jī)系數(shù)泛函自回歸模型的幾何遍歷性.pdf

1、分類號UDC密級編號中南大學(xué)CENTRALSOUTIIUNIVERSITY碩士學(xué)位論文論文題目一兩蒸隆扭基數(shù)丞魚自旦坦模型煎且何遍歷絲學(xué)科、專業(yè)趣疊造量數(shù)堡統(tǒng)啦研究生姓名導(dǎo)師姓名及專業(yè)技術(shù)職務(wù)唐盟曼一酆攙一_中熬攮～2006年11月ABSTRACTNonlineartimeseriesanalysisisamorecommonandmoreextensiveusingprospectthanthelinearoneAtpresent，i

2、thasbecomeanimportantresearchdirectioninthetimeseriesanalysistheoryOneof。theimportantmethodsofresearchingnonlineartimeseriesismodels’analysisRecently，manynonlineartimeseriesmodelshavebeenproposedandobtainedwidelyusedaswe

3、llButnowadaysinthewidelystudiedmodelsthedisturbanceisasin#ewhitenoiseThiskindofmodelshasobviouslimitthatiSthemodelsdon’tconsiderdisturbancecausedbyenvironmentsIntheactualsituationmedisturbalicetothesystemvarieswiththedif

4、ferentenvironmentsOntheotherhand，SOfarthestudyonnon—linetimeseriesisonthelengthofdelayisaconstantHoweverinfactitisnottruesometimesThetimeseriesmodelsstudiedbythismaster’thesisunderrandomenvironmentandrandomdelaysimulatet

5、hephenomenonthatdynamicsystemisinterferedbyrandomenvironmentmuchbetterthanpreviousonesAtthisrespectHouZhentingeta1prolessorsofInstituteofProbability&StatisticsofCentralSouthUIliversityhavecarriedcorrelativestudiesandanal

6、ysisTheauthoradoptstheirideas，andbyapplyingtheergodicitytheoryofMarkovchainsongeneralstatespace，discussedtwokindsoftheRandomCoe伍cientFunctionalAutoregressiveandobtainedsufficientconditiOIlSfortheirconvergenceundersomecon

7、ditionsTlliSthesisconsistsoffourchapters：ChapterlbrieflyintroducesthesituationsoftimeseriesanalysisChapter2，鵲basicknowledge，presentssomeergodicitytheoryofMarkovchainsongeneralstatespaceSOwecanuselaterChapter3，theauthorpr

8、oposesakindofRandomCoe伍cientFunctionalAutoregressivemodelunderrandomenvironmentandit’SgeometricergodicityisdiscussedChapter4，akindofRandomCoefficientFunctionalAutoregressivemodelwithrandomdelayisproposedandit’Sgeometrice