简介:Surfacequalityhasbeenoneofthekeyfactorsinfluencingtheongoingimprovementofthequalityofsteel.Therefore,itisurgenttoprovidemethodsforefficientsupervisionofsurfacedefects.Thispaperfirstexpressedthemainproblemsexistingindefectmanagementandthenfocusedonconstructingadataplatformofsurfacedefectmanagementusingamultidimensionaldatabase.Finally,someon-lineapplicationsoftheplatformatBaosteelweredemonstrated.Resultsshowthattheconstructedmultidimensionaldatabaseprovidesmorestructureddefectdata,andthusitissuitableforswiftandmulti-angleanalysisofthedefectdata.
简介:Alloyingelements,suchassiliconandmanganese,haveamajorimpactonthephasetransformationpointofsteel.Specifically,manganeseisanelementfortheexpansionandstabilityoftheausteniteregion,whilesiliconcanexpandandstabilizetheferriteregion.Phasetransformationoccursduringthehotrollingprocessforthesteelwithcertainsiliconcontent,whichleadstogreatchangesofthedeformationresistance,therebyaffectingtherollingstability.Consequently,abetterunderstandingofphasetransformationintherollingprocesswillcontributetotheenhancementofproductquality.Inthispaper,theon-linerollingdatawereprocessedbymeansoftheinversecalculationmethod.Inthismethod,thesteeldeformationresistancewithvarioussiliconandmanganesecontentswasobtainedandanalyzedtodeterminethedeformationbehaviorofthesteel,whichcanhelpimprovetheon-linecontrolmodelandenhancethesteelquality.
简介:Thesynthesis,compositionoptimization,VUVPhotoluminescence(PL)spectra,andopticalproperties,of(Y,Gd)(V,P)O4∶Eu3+phosphorswereinvestigatedbysynchrotronradiation.TheVUVPLEspectraandthecorrelationamongVUVPLintensity,λem,andEu3+,Gd3+,andP-contentwereestablished.ThePLEspectralstudiesshowedthat(Y,Gd)(V,P)O4∶Eu3+exhibitedsignificantabsorptionintheVUVrange.TheVUVPLintensitywasfoundtoenhancewithPO43-andGd3+-doping.Furthermore,thechromaticitycharacteristicsof(Y,Gd)(V,P)O4∶Eu3+werealsofoundtobe(0.6614,0.3286)andcomparedagainst(Y,Gd)BO3∶Eu3+asareference.Basedonthecharacterizationresults,wearecurrentlyimprovingandevaluatingthepotentialapplicationof(Y,Gd)(V,P)O4∶Eu3+asanewred-emittingPDPphosphor.
简介:目前,我国商业银行所面临的信用风险随着信贷业务的不断发展而逐步增加,如何对企业信用风险进行有效区分和管理,是商业银行亟待解决的问题.基于此,本文依据信用评估指标体系分别对Logistic回归模型、贝叶斯判别模型、支持向量机模型这三类模型进行了设计与构建,同时对三类模型分别进行实证分析和结果评价,从分类准确率和模型稳健性两方面对结果进行比较,作为进一步建立组合分类预测模型的基础.本文的研究成果,有利于推动我国商业银行信用风险定量度量方法的研究,从而有助于提高商业银行的风险控制水平,使得不良资产得以降低,在提高我国商业核心竞争力以及促进消费信贷市场的发展等方面有巨大的意义.