摘要
为评估三峡库区频发微震与周期性水流力对库区桥梁悬臂浇筑施工的合龙影响,采用有限元-正交试验设计-神经网络-蒙特卡罗模拟的联合施工风险评估方法对其进行分析.以正交试验设计法生成水流速度、水流高度、微震加速度的概率分布样本,耦合有限元模型与BP神经网络进行训练,建立各控制参数与中跨合龙中轴线偏移值之间的非线性映射关系,然后利用蒙特卡罗模拟随机生成影响参数,代入训练好的BP神经网络预测中轴线偏移误差风险.结果显示,水流力-高频微震耦合作用使合龙段中轴线偏移的风险概率已达到5.06%,库区大桥悬臂浇筑施工线型偏离风险需要引起足够的关注与防范.
In order to evaluate the closure effect of frequent microseisms and periodic water flow force on bridge cantilever construction in the Three Gorges Reservoir Area,a joint construction risk assessment method of finite element-orthogonal experimental design-neural network-Monte Carlo simulation was used to analyze the closure effect.The probability distribution samples of water flow velocity,water flow height and microseismic acceleration were generated by orthogonal experimental design method,and the finite element model and BP neural network were coupled for training.The nonlinear mapping relationship between each control parameter and the central axis offset value of the mid-span closure was established.Then Monte Carlo simulation was used to randomly generate influence parameters,which are substituted into trained BP neural network to predict the central axis offset error risk.The results show that the coupling effect of water flow force and high frequency microseismic makes the risk probability of axis deviation of closure section reach 5.06%,and the risk of line deviation in cantilever casting construction of reservoir bridge needs to be paid enough attention and prevented.
作者
付军
张涛
谢逸超
肖开乾
梁冠亭
马晓冬
FU Jun;ZHANG Tao;XIE Yichao;XIAO KaiQian;LIANG GuanTing;MA Xiaodong(School of Transportation,Wuhan University of Technology,Wuhan 430063,China;ubei Yichang Yiling District Highway Administration Bureau,Yichang 443100,China;Wuhan Municipal Construction Group Co.Ltd.,Wuhan 430056,China;Hubei Zixing Changjiang River Bridge Construction Development Co.,Ltd.,Zigui 443699,China)
出处
《武汉理工大学学报（交通科学与工程版）》
2020年第1期81-84,90,共5页
Journal of Wuhan University of Technology（Transportation Science & Engineering）
基金
湖北省交通科技项目资助(2017-2-16)。
