1、學位論文獨創(chuàng)性說明本人鄭重聲明：所呈交的學位論文是我個人在導師指導下進行的研究工作及其取得研究成果。盡我所知，除了文中加以標注和致謝的地方外，論文中不包含其他人或集體已經公開發(fā)表或撰寫過的研究成果，也不包含為獲得西安科技大學或其他教育機構的學位或證書所使用過的材料。與我一同工作的同志對本研究所做的任何貢獻均已在論文中做了明確的說明并表示了謝意。學位論文作者簽名日期：學位論文知識產權聲明書本人完全了解學校有關保護知識產權的規(guī)定，即：研究生

2、在校攻讀學位期間論文工作的知識產權單位屬于西安科技大學。學校有權保留并向國家有關部門或機構送交論文的復印件和電子版。本人允許論文被查閱和借閱。學?？梢詫⒈緦W位論文的全部或部分內容編入有關數據庫進行檢索，可以采用影印、縮印或掃描等復制手段保存和匯編本學位論文。同時本人保證，畢業(yè)后結合學位論文研究課題再撰寫的文章一律注明作者單位為西安科技大學。保密論文待解密后適用本聲明。學位論文作者簽名：指導教師簽名：年月日Subject:Theonlin

3、emonitingresearchonchangepointofregressionmodelsSpecialty:AppliedMathematicsName:ZhaoSha(Signature)Instruct:ZhaoGaochangJinHao(Signature)ABSTRACTThemainresearchofthispaperistheproblemofonlinemonitingpointintheregressionm

4、odelsfirstweaccdingtothetimeseriesbuildtheregressionmodelsusingtheleastsquaresmethodtoestimatetheparameterssowecangettheresidualexpression.Basedonthedifferentsampletoestimatetheparameterwedefinethegeneralresidualsbyusing

5、thebefemdatawhichisnonpollutingatthesametimewecandefinetherecursiveresidualsbyusingthebefe1i?datawhichhavebeendetected.Secondwecanaccdingtotheexpressionofthegeneralresidualsbuildtheonlinemonitingstatisticsofthemeanvarian

6、cewhichbasedonthegeneralresidualsatthesametimewecanaccdtotheexpressionoftherecursiveresidualsbuildtheonlinemonitingstatisticsofthemeanvariancewhichbasedontherecursiveresiduals.Fromthetheem1to10wecanobtainedthesefourstati

7、sticsexistadifferentlimitunderthenullhypothesisthelimittendstoinfinityunderthealternativehypothesis.Sowecanprovethatthesefourstatisticsareeffectiveinthey.FfurtherillustratetheeffectivenessofthesestaticswecanusetheMonteCa

8、rlomethodtosimulatetheexperimentssowecanobtainedsomecriticalvalueslevelofexperiencepotentialtheaveragerunlengthunderthealternativehypothesisofthesefourstatistics.Throughtheanalysisofsimulationresultswecanobtainedthefollo

9、wingconclusions:thepotentialofthetestaffectedbythelocationofthepointthewindowwidthparameterhgamathetypeofthechangepoint.Theresultofthevarianceonlinemonitingstatisticsbasedontherecursiveresidualisnotveryobviousbuttheother

10、threestatisticsareverysensitiveIndertogetthehigherpotentialsmalleraveragerunlengthweshouldusethemeanonlinestatisticsbasedontherecursiveresidualwhichthewindowwidthparameteris0.2h?totestthemeanchangepointIndertotestthevari