Page 68 - 2022年第53卷第10期
P. 68
[26] BUF.Anefficientfuzzyc - meansapproachbasedoncanonicalpolyadicdecompositionforclusteringbigdatainIoT
[J].FutureGenerationComputerSystems,2018,88:675 - 682.
[27] PALNR,BEZDEKJC.Onclustervalidityforthefuzzyc - meansmodel[J].IEEETransactionsonFuzzySys
tems,2002,3(3):370 - 379.
[28] DINGY,FUX.Kernel - basedfuzzyc - meansclusteringalgorithm basedongeneticalgorithm[J].Neurocomput
ing,2016,188:233 - 238.
[29] MANIMALAK,DAVIDIG,SELVIK.AnoveldataselectiontechniqueusingfuzzyC - meansclusteringtoen
hanceSVM- basedpowerqualityclassification [J].SoftComputing,2015,19(11):3123 - 3144.
[30] 朱素霞,祖宏亮,孙广路.一种基于空间信息的 FSICM 图像分割算法 [J].哈尔滨理工大学学报,2020,
25(4):101 - 108.
[31] ATANASSOVK.Intuitionisticfuzzysets[J].InternationalJournalBioautomation,2016,20:1.
[32] SUGENOM.FuzzyMeasuresandFuzzyIntegrals:ASurvey[M].MorganKaufmann,1993.
[33] 王娜,李霞.一种新的改进 Canny边缘检测算法[J].深圳大学学报,2005(2):149 - 153.
[34] KEMENYJM,DEVGANA,HAGAMANRM.Analysisofrockfragmentationusingdigitalimageprocessing[J].
JournalofGeotechnicalandGeoenvironmentalEngineering ,1993,119(7):1144 - 1160.
[35] KEMENY,JOHNM,AUTHORA.Practicaltechniquefordeterminingthesizedistributionofblastedbenches,
wastedumpsandheapleachsites[J].MiningEngineering,1994,46(11):1281 - 1284.
[36] HARDYAJ,RYANTM,KEMENYJM.Blocksizedistributionofinsiturockmassesusingdigitalimagepro
cessingofdrillcore [J].InternationalJournalofRockMechanicsandMiningSciences&GeomechanicsAbstracts,
1997,34(2):303 - 307.
Intelligentdetectionmethodformaterialqualificationofearth - rock
dam basedondigitalimageprocessing
1
2
2
1
3
ZHAOYufei,LIUBiao,WANGYi,MENGLiang,LIUBiwang
(1.ChinaStateKeyLaboratoryofSimulationandRegulationofWaterCycleinRiverBasin,ChinaInstituteof
WaterResourcesandHydropowerResearch ,Beijing 100038,China;
2.SinohydroBureau8Co.,Ltd.,Changsha 410007,China;3.SinohydroBureau6Co.,Ltd.,Shenyang 110179,China)
Abstract:Thequalificationtestingofearth - rockdammaterialsisusuallyrealizedbyjudgingwhetherthegradation
characteristicparametersobtainedfromon - sitescreeningtestmeetthedesignrequirements.However ,themethod
ofobtainingthegradationcharacteristicparametersthroughthetesthassomeshortcomings,suchaslowsampling
rate ,cumbersomeoperationprocessandpoorintelligentperception,resultinginpoorrepresentativenessofthetes
tingresults.Inordertoimprovetheintelligentdetectionofdammaterialgradationparameters,relyingontheima
gesandgradationdataatthetestlocationofonepumpedstoragepowerstationinLiaoningProvince ,theintuitionis
ticfuzzy C - meansclusteringalgorithm fusedwithspatialinformation(SIFCM)isusedtosegmenttheimageof
earth - rockdam materials.Next,the3D volumereconstructionofearth - rockdam materialisachievedbythe
equivalentellipsoidalvolumemethod.Thenthegradationcharacteristiccurveofdammaterialunderrealconditions
isobtainedthroughthegradationcorrectionmodelbasedonBPneuralnetwork.Finally ,fourevaluationindexesof
dam materialqualificationareobtained : maximum particlesize, P5content, curvaturecoefficientC c , and
unevencoefficient C u .Thepracticalengineeringapplicationshowsthattheintelligentidentificationandcorrection
modelofdammaterialgradationcharacteristicsbasedontheSIFCM_BPalgorithmestablishedinthispaperhashigh
identificationaccuracy.Themethodinthispaperprovidesanimportantsupportfortherapididentificationofdam
materialqualificationbeforethecompactionconstructionandthereal - timeevaluationofdam materialcompaction
characteristicsduringconstruction.
Keywords:materialgradationdetectionofearth - rockdam;digitalimageprocessing;SIFCM algorithm;gradation
correctionmodel;dammaterialqualificationdetection
(责任编辑:李 娜)
— 1 2 6 —
0