摘要
特大断面隧道开挖会大范围解除围岩表面应力,在上部岩体自重应力及附加应力的共同作用下,隧道拱顶会发生明暕沉降,在某些软岩隧道中已观察到了超过100 mm的沉降量。拱顶沉降量持续增大,隧道就有可能发生坍塌破坏。在隧道开挖工程中,受条件限制,常会出现支护无法及时施工的情况,需要对隧道稳定性进行分析预测。现以牛寨山隧道开挖工程南线出口段的监测结果为例,采用BP神经网络方法预测其沉降量的时程变化。假设隧道相向开挖过程中,在贯通前对拱顶沉降互不影响,且两条隧道之间互不影响。分析过程中输入层参数选择了围岩级别、埋深、距离掌子面和二衬的长度等,隐含层设置为1层,节点数为9,隐层传递函数选择tansig,输入输出层传递函数选择purelin。在采用该模型进行预测分析前,首先已开挖监测点数据作为训练样本,采用后开挖点的值作为样本,验证了模型的可靠性。最终分析结果表明,采用文中方法能够较为可靠地预测拱顶沉降。其结果可以作为隧道安全预测的依据。
The excavation of super large sectional tunnel will relieve the surface stress of surrounding rock in wide range.The tunnel vault will cause the obvious settlement under the joint action of the self-weight stress and additional stress of the upper rock.The settlement volume exceeding 100 mm has been observed in some soft rock tunnels.The tunnel will possibly collapse and be damaged if the vault settlement volume continuously increases.It is often unable to construct the support because of the limited conditions in the process of tunnel excavation.It is required to analyze and forecast the stability of tunnel.Taking the monitoring result at the exit section of the south line of Niuzhai Mountain Tunnel Excavation Project as an example,the time course changes of its settlement volume are forecast by BP nervous network method.It is supposed that the vault settlements have no influence before the breakthrough in the process of tunnel opposite excavation,and there is no influence between two tunnels.In the analysis process,the layer parameters are inputted to select the rock level,depth,distance from the tunnel face and two- line length.The hidden layer is set up by one layer.The node number is 9.The transfer function of hidden layer is inputted to select transig.The transfer function of output layer is inputted to select purelin.Before the forecast analysis by using this model,it is firsdy to take the data of the excavated monitoring point as the training sample,and the value of excavation point is taken as the example after used to validate the reliability of the model.The final analysis result shows that the use of the described method can more reliably forecast the vault settlement.Its result can be referred as the basis to forecast the tunnel safety.
出处
《城市道桥与防洪》
2015年第1期145-150,15-16,共6页
Urban Roads Bridges & Flood Control
关键词
特大断面隧道
拱顶沉降
BP神经网络
时程曲线
super large sectional tunnel
vault settlement
BP nervous network
time course curve
