欢迎来到天天文库
浏览记录
ID:52174889
大小:666.59 KB
页数:5页
时间:2020-03-23
《基于BP网络的深部软岩巷道围岩力学参数反演研究.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、第33卷第5期矿冶工程V01.33№52013年1O月MININGANDMETALLURGICALENGINEERING0ctober2013基于BP网络的深部软岩巷道围岩力学参数反演研究①夏海燕,付建军,时凯(1.湖北工程学院城市建设学院,湖jE武汉410073;2.中国水电顾问集团中南勘测设计研究院,湖南长沙4100142;3.中国科学院武汉岩土力学研究所,湖北武汉430071)摘要:为探索高地应力下深部软岩巷道围岩应力场及位移场的分布情况,建立了简化的应变软化模型,分析了此模型的关键参数;采用均匀设计法设计了数值计算中待分析的参数方案,且将各方案参数代入FLAC软
2、件中进行了数值仿真分析;再将各工况下监测点计算的相对位移与岩体力学参数代人BP网络进行训练、学习,并将监测点位移相对监值输入训练的BP网络,获得参数反演值。以淮南矿业集团顾桥煤矿轨道巷为工程背景,进行了巷道深部位移监测,根据监测结果反演了简化应变软化模型的关键参数,经正分析对比验证了力学参数反演结果的准确度、可靠性,反演结果为后续工程数值计算奠定了基础。关键词:软岩巷道;围岩;高应力;应变软化模型;均匀设计;BP网络中图分类号:TU443文献标识码:Adoi:10.3969/j.issn.0253-6099.2013.05.006文章编号:0253-6099(2013)
3、05—0025—05StudyonParametricInversionofRockMechanicsforSurroundingRockinDeepSoftRockRoadwaybyBPNeuralNetworkMethodXIAHai.yan,FUJian-jun,SHIKai(1.CollegeofUrbanConstruction,HubeiEngineeringUniversity,Wuhan410073,Hubei,China;2.HYDROCHINAZhongnanEngineeringCorporation,Changsha410014,Hunan,Ch
4、ina;3.InstituteofRockandSoilMechanics,ChineseAcademyofSciences,Wuhan430071,Hubei,China)Abstract:Inordertoexplorethefieldofstressanddisplacementfieldofthesurroundingrockindeepsoftrockroadwayundertheconditionofhighstressandlowstrength,asimplifiedstrain-softeningmodelwasestablishedandkeypar
5、ametersofthismodelwereanalyzed.Theparameterprogramtobeanalyzedinnumericalcomputationwasdesignedwithauniformdesignmethod,thenofwhichallparametersweresubstitutedintoFLAC。。softwarefornumericalsimulationanalysis.Afterthat,therelativedisplacementandrockmassmechanicalparametersofeachmonitoring
6、pointunderdifferentworkingconditionsweresubstitutedintoBPneuralnetwork,fortrainingandstudy,thendisplacementrelativetomonitoringwasinputintothetrainedBPneuralnetworktoobtaintherealparameters.WiththesoftrockroadwayinGuqiaoCoalMineofHuainanMiningGroupasanengineeringpractice,keyparametersint
7、hesimplifiedstrain-softeningmodelwereprocessedbyparametricinversionbasedonin-situmonitoringofthedeepdisplacementofsuroundingrock.Theresuhhasbeenverifiedtobeaccurateandreliableafteranalysisandcomparison,layingthegroundworkforthefollowingengineeringnumericalanalysis.Keyword
此文档下载收益归作者所有