日徑流預(yù)報(bào)方法研究.pdf

1、華中科技大學(xué)碩士學(xué)位論文日徑流預(yù)報(bào)方法研究姓名：周娜申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：水利水電工程指導(dǎo)教師：張勇傳李承軍2003.5.10華中科技大學(xué)碩士學(xué)位論文AbstractInthispaperthemodelsareconstructedtoforecastthevalueofdallyrunoffItsresultnotonlyprovidestheevidenceofdecisionmaking，butalsoimpactsgreat

2、lyoneconomicaloperationofhydropowersystem，navigation，floodcontrollingandSOonThetheoryanddevelopmentofforecastingisrefiewed，andthecorrelativemethodsofrunoffpredictionarediscusseddeeplyassociatedwitIltheprojcctof“VirtualCa

3、scadedScheduleofQingjiangHydropowerSystem”ThecorrectandintegrateddataisgainedafterdisposingthehistoricaldatatobefitforusinginmodelsUsedthedataofrunoffandrainfall，amodelofrainfallrunoffisbuiltinChapter2Traditionaltheoryof

4、rainfallrunoffhasbeengrownupButbecauseofthelackofholdingdatait’SverydifficdttobebuiltAtthesametime，traditionalmodeltakestoomuchtimetobeconstructedandcomputedanddoesn’tsuittobereplantedUsingtheadvanceofcompmcracorrelation

5、modelofrunoffandrainfallisbuilttodescribetheflushwaterperiodThemaintheoryoftimeseries，especiallythelinearstationarytimeseries，hasbeengrownupIthasbeenusedtocontrolandforecastinvariousfieldsInthispaperthecharactem，correlat

6、iveanalysismethodsandmodelingprocessofARMAseriesareintroducedARmodelisgivenaccordingtocorrelativeanalysisbasedonavmlabledatatoforecasttherunoffindroughtperiodBypractice，goodmsdtisattainedThetwoabovemodelsonlycanmakegoodr

7、esultinsomepartsofannualrunoffDecauseitiSdi伍culttodescribearunoffcoursewithlinearsysteminthenatureThenon—linearsystemssuchasartificialneuralnetworkgeneticalgorithmareintroducedBPmodelwitllthethree—layerandGAmodelaI℃found

8、edtosimulatetheall—yearrunoffMoreover’theunitedmodelcombinesthetwoalgorithmstoforecastAfterm她someusefulresearchtooptimizethenetworkconstructandalgorithm，theunitedmodelsateusedtopredicttheannualrunoffanditisfeasiblebytest