欢迎来到天天文库
浏览记录
ID:31961949
大小:7.04 MB
页数:72页
时间:2019-01-29
《河道洪水演进马斯京根模型参数与最优参数估计方法-研究》由会员上传分享,免费在线阅读,更多相关内容在应用文档-天天文库。
1、万方数据minimizetheerrorsumofsquares.Then,tosolvebythenonlinearprogramming,itisconcludedthattheerrorsumofsquaresthantrialanderrormethodreducebigger,thecorrelationcoefficientincreasing.Itsuggeststhatthemethodusingtheerrorsumofsquaresminimumofcalculateddischargeandmeasureddischargeast
2、hecriteriontooptimizeflowroutingcoefficientCo,ClandC2,thenreverseKandX,isfeasible,butitisnottheoreticalbasisfortheglobaloptimalsolution;Finally,withtheswarmalgorithmofcalc。ulationsimpleandthecontrolparametersless,goodconvergence,strongrobustness,strongglobalsearchabilityandother
3、characteristicstooptimizeMuskingummodelparameters:startwithanonlinearfunctiontoverifytheextremaloptimizationabilityofswarm}algorithm,thenusedtoobtainoptimalsolution,theresuksshowthattheerrorsumofsquaresofcalculateddischargeandmeasureddischargecomparedwith;thenonlinearprogramming
4、methodreduceinT~ia;ndongstation,thecorrelation,coefficientincreasing.Andthentotestthesuperiorityofthealgorithm,÷comparingwiththecommonlyusedparticleswa.rmalgorithmoptimizingthe}results,Swarmalgorithm,optimizationtogetthe;errorjlower,correlationcoefficientslightlylargersumofsquar
5、esslightlythanparticleswarmoptimization.Butunderthesamenumberofiterations,theswarmalgorithmconvergesfaster,moreclosetotheglobaloptimalvalue.SotheMuskingummodelparametersoptimizationbasedonswarmalgorithmhashigheraccuracyandscientificity.Swarmalgorithmoptimizationresultscanbeuseda
6、sMuskingummodelparametervaluesoftheBaise·111万方数据Tiandongriver,thatmethodoptimizationshows:withtheincreaseofthemagnitudeofthepeakflowoffivefloodinBaisehydrologicalstation,Kvaluesreducebiggeraccordingly,SOtocarryonoptimalfittingoftheBaisestation’SfloodpeakflowandK.Inthefuture,acco
7、rdingtothefittingformulatocalculatethecorrespondingKvaluethroughthepeakflowintheriverfloodroutingissuggest,andXvalueschangerelativelystable,SOaveragevalue0.24ofthefivefloodcalculatedresultscanbeusedasvalue工.KEYWORDS:Muskingummodel;Intervalinflow;Trialanderrormethod;Nonlinearprog
8、rammingmethod;Swarmalgorithm;Globaloptimalsolut
此文档下载收益归作者所有