Page 49 - 2024年第55卷第7期
P. 49

SimulationmethodsofbankfulldischargeintheLowerYellowRiver
                                                                       1,2,3
                                                          1,2,3
                                             1,2,3
                                 WUBaosheng     ,SHENYi ,MAZipu           ,ZHENGShan   4
                          (1.StateKeyLaboratoryofHydroscienceandEngineering,TsinghuaUniversity,Beijing 100084,China;
                     2.KeyLaboratoryofHydrosphereSciencesoftheMinistryofWaterResources,TsinghuaUniversity,Beijing 100084,China;
                                3.DepartmentofHydraulicEngineering,TsinghuaUniversity,Beijing 100084,China;
                                    4.SchoolofCivilEngineering,TianjinUniversity,Tianjin 300072,China)

                  Abstract:UnderthebackgroundofwaterandsedimentvariationintheYellowRiverBasin,accuratecalculation
                  ofthebankfulldischargeofthelowerYellowRiveriscrucialforpredictingthetrendofrivererosionandsedimenta
                  tionandcontrollingthesuspendedriver.Onthebasisofcomparingandanalyzingmultipleexistingmethodsforcal
                  culatingthebankfulldischargeinthelowerYellowRiver,thisarticlefirstestablishesadelayedresponsemodelfor
                  bankfulldischargeconsideringtheeffectofnon - floodseasonwaterandsedimentconditions.Then ,asimulation
                  methodforbankfulldischargebasedondelayedresponsemodelandenergymethodisproposed.Theresultsshow
                  that :(1)thesignificantchangeshaveoccurredinnon - floodseasonwaterandsedimentconditionsaftertheopera
                  tionoftheXiaolangdireservoir.Amongthethreemethodsproposedinpreviousstudiesforcalculatingthebankfull
                  dischargeinthelowerYellowRiver,theslidingaveragemethod,delayedresponsemodel,andenergymethodare
                  allunabletoeffectivelysimulatethebankfulldischargeinthelowerYellowRiverafter2000.(2)Adelayedre
                  sponsemodelconsideringtheeffectofnon - floodseasonwaterandsedimentconditionshasbeenestablishedtore
                  flecttheimpactofnon - floodseasonwaterandsedimentconditionssincetheoperationofXiaolangdiReservoir.Al
                  thoughthesimulationaccuracyofbankfulldischargehasbeenimproved , someparametersinthecalculation
                  formulaforthebalancevalueofbankfulldischargehavesignificantdifferencesandthestructuralformisunstable.
                  (3)Drawingontheadvantagesoftheenergymethod,theformulaforcalculatingtheequilibriumvalueofbankfull
                  dischargehasbeenimproved.Theimprovedbankfulldischargecalculationmethodcannotonlyreflecttheimpactof
                  annualincomingwaterandsediment ,butalsosimultaneouslyreflecttheeffectofbedresistancerepresentedbythe
                  medianparticlesizeofbedsediment.Thenewmethodnotonlyhasacertaintheoreticalbasis ,butalsoimproves
                  thesimulationaccuracyofbankfulldischarge.
                  Keywords:LowerYellowRiver;bankfulldischarge;waterandsediment;mediansizeofbedsediment;delayed
                  response ;energymethod
                                                                                    (责任编辑:李福田)


              (上接第 790页)
                             Anearlywarningmethodforpotentialcollisionriskofdam surface
                               constructionmachinerybasedonlargescenevideosurveillance
                                                  1
                                    1
                                                                                                  3
                                                                                     1
                                                                    1,2
                      ZENGTuocheng,WANGJiajun,ZHONGDenghua ,ZHANGYunuo,KANGDong
                 (1.StateKeyLaboratoryofHydraulicEngineeringIntelligentConstructionandOperation,TianjinUniversity,Tianjin 300350,China;
                         2.CollegeofWaterResourcesandCivilEngineering,ChinaAgriculturalUniversity,Beijing 100083,China;
                                  3.YalongRiverHydropowerDevelopmentCo.,LTD,Chengdu 610051,China)
                  Abstract:Basedonlarge - scenevideomonitoring,itisimportanttorealizethepotentialcollisionriskwarningof
                  damconstructionmachinerytoensurethesafetyofdamsurfaceconstruction.However ,atpresent,itmainlyre
                  liesonmanualexperiencetojudgethepotentialcollisionriskofdam surfaceconstructionmachinery ,whichis
                  pronetoproblemssuchasomissionandmisjudgment.Therefore ,thispaperproposesanearlywarningmethodfor
                  potentialcollisionriskofdam surfaceconstructionmachinerybasedonlarge - scenevideosurveillance.First,
                  basedonTrajectron ++algorithm,thetrajectorypredictionofmulti - typeconstructionmachinerywithlargequanti
                  tyinthefutureperiodoftimeinthelarge - scenevideosurveillanceofthedamsurfaceisrealizedthroughtransfer
                  learning ;then,thedriveproximitytimeandmaximumcrowdednessofmachineryareproposedasthequantitative
                  indexesofthepotentialcollisionriskoftheconstructionmachineryonthedamsurface ,andthefuzzymembership
                  functionofpotentialcollisionriskgradedwarningunderdifferentdrivingspeedsofconstructionmachineryisestab
                  lishedbasedonthefuzzyrules ;finally,thewarningresultsofthetwoindicatorsarefusedwithevidencefusion
                  theorytoobtainthefinalwarninglevel.Theproposedmethodistestedonthelarge - scenevideosurveillanceofthe
                  Lianghekoudam construction site.Theaveragedisplacementerrorand thefinaldisplacementerrorofthe
                  trajectorypredictionoftheconstructionmachineryonthedamsurfaceinthenext6secondsare1.17and2.36m ,
                  respectively,andthefuzzy - evidencefusion - basedpotentialcollisionriskgradingwarningresultsoftheconstruc
                  tionmachineryprovideanautomatedandintelligentanalysismethodfortheconstructionsafetyofthedamsurface.
                  Keywords:constructionmachinerysafety;large - scenevideosurveillance;trajectoryprediction;fuzzy - evidence
                  fusion ;gradedwarning
                                                                                    (责任编辑:韩 昆)

                                                                                                —  8 0 1 —
   44   45   46   47   48   49   50   51   52   53   54