Page 202 - 《水资源与水工程学报》2023年第5期
P. 202
! 34 " ! 5 # & ' ( ) & * + , - Vol.34No.5
2023 $ 10 % JournalofWaterResources&WaterEngineering Oct.,2023
DOI:10.11705/j.issn.1672-643X.2023.05.23
'P GAMLSS-GLOoB"øU
,tKh$ò
1 1 1 1 1 1 1 1 2
JwÒ , Ó Ô , ÕÖÅ , J×Ø , N o , v:Ù , Ú , ÛÜK , ÝÆÆ
(1. `$T3CçK¹ÈZ;.efJCçW, ±Q g 665005;2. `$T3CçK¹ÈZ;., ±Q H½ 650214)
@ A: DÃú]ÔèJ, l3CÒ、 ¿CêYf|PQ, Àkú]Ô|Kg\c, àjx¯]ÔJ
º», _ú!Åâ)õ|þÿ。20, DêjGÄj, ø¹ Óz%^c¿C[\r
]Ô6W¶rVCÒ|ª«; }~]Ô6JA5Å[¿C|ôàPQ, gLÅÚ(GAMLSS-GLO),
À¿CPQrú]ÔJ4|îTa£, ýW]Ô CÒª«ÓÚ, IJú]ÔJ|Å£a
£; Ëa, ýȸ¹efJPú|]ÔJ。º»¼½: Ä]Ô|CÒ、 ¿CQ¸¼z
x|¹S, Ðn%]ÔÀIJ©6。~()]Ô6ora£, Èdú|é¯HÃgv«
!。
BCD: ú]Ô6; ª«¦§;GAMLSS ; Óz%; a£IJ
EFGHI:TV641 JK7LM:A JNOI:1672643X(2023)05019809
Seepagemonitoringmodelofearth-rockdam basedon
GAMLSS -GLOandrandom forestalgorithm
1
1
1
1
1
ZHAOMingzhe,FANMu,SHENLifeng,ZHAOPeishuang,MAJun,
1
2
1
1
TANGXiaosong,LUOHang,SONGBoxu,NIEBingbing
(1.NuozhaduPowerPlantofHuanengLancangRiverHydropowerInc.,Puer665005,China;
2.HuanengLancangRiverHydropowerInc.,Kuming650214,China)
Abstract:Thekeyproblemsoftheseepageflowmonitoringofearth-rockdamsarehowtoclarifythein
fluenceofwaterlevel,rainfallandotherfactors,howtosimulatetherealseepageflowandevaluatethe
resultsoftheseepagemonitoringbasedontheprototypeobservation.Inviewofthis ,aregressionmodel
ofseepageflowandupstreamanddownstreamwaterlevelsisestablishedundertheconditionofnorainfall
usingtherandomforestalgorithmandtheprinciplesofmathematicalstatistics.Regardingtothecompre
hensiveinfluenceoftheseepageflowontheaccumulatedrainfallintheearlierperiod ,theGeneralized
AdditiveModelsforLocation,ScaleandShapeofGeneralizedLogisticDistribution(GAMLSS-GLO)is
introducedtosimulatethefluctuationintervalsofthemonitoringvaluesofseepageflowundertheinflu
enceofrainfall.Then ,GAMLSS-GLOissuperimposedwiththeseepageflow-waterlevelregression
modeltopredictthereliableintervalsoftheseepageflow.Finally ,themethodisappliedtotheseepage
monitoringofthecorewallearth-rockdaminNuozhaduReservoir.Theresultsshowthattheproposed
methodshowsgoodapplicabilitytothewaterlevelandrainfallresponseofseepageflow ,itcansignifi
cantlyimprovethequalityofseepagesimulationprediction ,andsolvetheconfidenceintervalsofseepage
flow,whichisconducivetojudgingtheoperationconditionandsafetymonitoringofearth-rockdams.
Keywords:seepageflowofearth-rockdam;regressionanalysis;GAMLSSmodel;randomforestalgo
rithm;intervalprediction
!"#$:20221129; %$:20230307
'()*:“ ” ./ef+æÄÙ-4(2022YFC3005401); ±Q[ef+æÄÙ-4(202203AA080009)
+,-.: ½²(1992—), N, ±QH½(, ;<, HI>, +J]yC5«CDE。
/0+,: ö ÷(1985—), 7, ±QÏU(, x<, HI>, +J]ykú!J。