基于粗糙神經網絡的番茄灰霉病預警模型構建與實現.pdf

1、獨創(chuàng)性聲明本人聲明所呈交的論文是我個人在導師指導下進行的研究工作及取得的研究成果。盡我所知，除了文中特別加以標注和致謝的地方外，論文中不包含其他人已經發(fā)表或撰寫過的研究成果，也不包含為獲得寧夏大學或其它教育機構的學位或證書而使用過的材料。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作了明確的說明并表示了謝意。研究生虢墑住帆釤／中年夕脅日關于論文使用授權的說明本人完全了解寧夏大學有關保留、使用學位論文的規(guī)定，即：學校有權保留送交論

2、文的復印件和磁盤，允許論文被查閱和借閱，可以采用影印、縮印或掃描等復制手段保存、匯編學位論文。同意寧夏大學可以用不同方式在不同媒體上發(fā)表、傳播學位論文的全部或部分內容。(保密的學位論文在解密后應遵守此協(xié)議)時間：糾r年孓月弘日時間：p，7，年，月聲日儀砂拭r互，d名名簽塹生師究引研AbstractThefacilityagricultureisevergrowingasashareofatotalagriculturalareaandi

3、tisaneffectiveprojecttosolvetheproblemthatmorepeopleandlesslandrestrictssustainabledevelopmentHowever，thedevelopmentoffacilityagricultureisfacedwithadifficultproblemthattheincreasinglyserioussituationoffacilityagricultur

4、ediseasesandinsectAtpresentthemonitoringdatacollectionmethodofthediseaseinformationmainlyreliesonfieldinvestigationcarriedoutbytheplantprotectionstaffs，samplingandanalysis，SOthedatasizeisnotenoughanddirectlyaffectstheacc

5、uracy，timelinessandeffectivenessoftheearlywarninganalysisAccordingtovariousenvironmentalparametersofthegreenhousetomatoeswhich盯erealtimegatheredthroughtheIntemetofthingstechnology，thetomatoBotrytiscinereaearlywarningmode

6、lbasedonroughneuralnetworkisbuiltanditCananalyzetherealtimegathereddataandthenforecasttomatoBotrytiscinereaThekeyresearchcontentsofthepaperareasfollows：1Inthegreenhouseenvironmentvarietyofrealtimedataisgatheredthroughvar

7、iouskindsofsensorsinperceptionlayeroftheIntemetofthingsanditmainlyincludesthetemperatureandhumidityoftheairandsoil，lightandcarbondioxideconcentrationwhichisstoredintemporarydatabasesofthecloudplatformandHadoopcluster2The

8、earlywarningmodelconstructionoftomatoBotrytiscinereaThepaperfirstlystudiestheeffectthatthedoubleenvironmentalfactorsimpactontomatoBotrytiscinereaandgetstheresultthatthereexistscertainlimitationsandthenmakesfurtherresearc

9、hbasedonmultifactorsBecausethenumberofthefactorismore，wecombinerespectivelytraditionalBPneuralnetworkandimprovedBPneuralnetworkandroughsettheoryThroughtheinstanceanalysis，theresultsshowthattheprecisionandtrainingtimeofth

10、eearlywarningmodelbasedonroughneuralnetworkwhichisprovidedinthepaperbothhavecertainenhancementTheestablishmentideaofthemodelCanprovidebeneficialreferencesfortheresearchoftheearlywarningofdiseaseandinsect3Theresearchandde

11、velopmentoftheearlywarningsystemsoftomatoBotrytiscinereaBasedonearlywarningmodelwhichhasbeenbuilt3蝌八language，MyEelipsedevelopmentenvironmentandSpringMVCarchitecture，theearlywarningsystemisdevelopedTherealtimedataisstored

12、inMySqlandHadoopclusterJSandJQueryandotherrelatedtechnologiesareusedforfrontpagedisplayBasedontheresearchthepapertakestomatoBotrytiscinereaasexampleanddevelopstomatoBotrytiscinereaearlywarningsystembasedonroughneuralnetw

13、orkThesystemispartoftheinfrastructurecloudservicesintheSAASlayeroftheagriculturalcloudserviceplatformofagriculturalcitywhichisconstructedbynationalmodemagriculturalcityforscienceandtechnologyAnditisdeployedandrunsintheag