基于信息熵理论的水文站网评价优化分析

《基于信息熵理论的水文站网评价优化分析》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、AbstractAsthecartierofhydrologicinformation，Hydrologicnetworklayoutdirectlyaffectstherepresentativenessandresearchvalueofhydrologicaldataprovidedbythehydrologicnetworkontheanalysisofinherentlawandcharacteristicsforbasin．However,theproblemof’’datafullybutlackofinform

2、ation”appears．Alargeamountofmanpowerandresourcesisputintotheconstructionofhydrologicnetwork，butthehydrologicdataproducesredundantinformation，whichhasbroughtinconvenienceandinterferencetotheresearchandapplication．Sowecanoptimizethehydrologicnetworktogetthemaximalinfo

3、rmation，whichcannotonlysaveresourceandinvestmentforhydrologicalconstruction，butalsoprovidescientificandsystematichydrologicaldataforhydrologyandwaterresourcesresearch．Thetraditionalnetworkoptimizationmethodhascertainlimitsanddisadvantages，andthispaperintroducedthein

4、formationEntropywhichisameasureoftheuncertaintyorinformationofsystem．Asitbasedoninformationtheory，informationentropymethodisdifferfromtraditionalmethodofnetworkoptimization．Itcanquantifytheamountofinformationprovidedbyasiteandtransmitedbetweensitesfromtheangleofsyst

5、ematicnessandintegrity,andestablishevaluationmodelandnetworkoptimizationmodeltoevaluateandoptimizethelayoutofthenetwork．Thispaperresearchsandanalysisthenetworkevaluationandoptimizationmodelbasedoninformationentropycombiningwiththeinstance，tryingtoprovidesanewfeasibl

6、enetworkoptimizationmethodforriverbasin．Themainresearchcontentsandresultsareasfollows：(1)Basedontherelatedbasicinformationentropytheory,andaccordingtothecharaeteristiesofthewatershedhydrologicnetworkanddata，whichisthetransferability,correlationandattenuationbetweens

7、itehydrologyinformation，thispaperdemonstratethereasonablenessandfeasibilityofhydrometricnetworkevaluationandoptimizationmethodbasedoninformationentropy,tolaythetheoreticalfoundationforfurtherexploration．(2)OnthebasisoftWOexistingmodelswhichisusedtodescribethelawofin

8、formationtransmissionbetweensites：transinformation—distancemodel(T-DModel)andcorrelation-distancemodel(C—DModel)，Thispaperputforwardtwonew