岩壁吊车梁硐室围岩力学参数反分析研究  

Back-Analysis of Surrounding Rock Mechanical Parameters for Cavern of Crane Beam

作　　者：阮永芬[1] 张永彬[1] 李仕胜 刘永芳 

机构地区：[1]昆明理工大学建筑工程学院,云南昆明650500 [2]中国水利顾问集团昆明勘测设计研究院科学研究分院,云南昆明650033

出　　处：《昆明理工大学学报（自然科学版）》2011年第6期35-40,共6页Journal of Kunming University of Science and Technology(Natural Science)

基　　金：云南省中青年学术技术带头人培养资助项目(2005PY01-36);云南省技术合同项目(20060023)

摘　　要：建立漫湾电站二期地下主厂房岩壁吊车梁硐室的三维有限元计算模型,对硐室围岩的力学参数和吊车荷载进行正交组合后,计算出各种组合情况下的围岩位移,对参数敏感性进行分析.把敏感性参数组合时计算出的位移值,作为BP神经网络训练的样本空间对模型进行训练,把监测值输入训练后的模型中反算出围岩力学参数,最后输入参数进行变形验算.即把三维有限元分析方法和神经网络方法联合应用,通过正分析-反分析-正分析的分析方法,计算出围岩的力学参数.对理论分析的结果和实际结果进行比较,证明了所采用的分析方法和分析模型是正确可行的,能为设计人员有效解决实际工程中的力学参数的准确取值问题,提供一种分析方法.Calculation of displacement of surrounding rock and sensitivity analysis of parameters are carried out by establishing a three dimensional finite element calculation model of crane beam of cavern of the second phase underground project in Manwan power station and by orthogonally combining mechanical parameters of surrounding rock and crane load.The calculation displacement with sentivity combined parameters is taken as sample space of BP neural network to train model.The monitored value is put in to trained BP neural network model for the back analysis of the mechanical parameters of surrounding rock.Finally,the input of parameters is used for computation checking.Through the combination of three dimensional finite element analysis method with BP neural network method,mechanical parameters are obtained by positive analysis-back analysis-positive analysis.The comparison between theoretical and actual results show that the analysis method and model is feasible,which can provide a good analysis method for designer to get accurate mechanical parameter.

关 键 词：反分析 BP神经网络 有限元 力学参数 岩壁吊车梁 

分 类 号：TU375.1[建筑科学—结构工程]