利用多点位移计实测值反演隧道围岩力学参数

Back Analysis for Surrounding Rock's Mechanical Parameters of Highway Tunnel by Using Measured Values from Multipoint Displacement Meter

结合哈尔滨天恒山隧道的多点位移计的测量工作,利用MATLAB BP神经网络工具箱和AN-SYS10.0,基于结构分析有限元法,由正演分析建立了不同围岩力学参数与隧道围岩变形关系的人工神经网络训练样本。结合施工过程中量测到的隧道围岩的多点位移计数值,通过网络仿真得到了相应的围岩力学参数。

Combining with the multipoint displacement meter＇ s measurements for Tianhengshan Tunnel of Ha＇erbin, by using MATLAB BP neural network toolbox and ANSYS 10.0, and based on the finite element method for structural analysis, the artificial neural network training samples for the relationship between the different mechanical parameters of surrounding rock and the surrounding rock＇ s deformation of tunnel are set up by positive direction analysis. And then, combining with the measured values of the tunnel＇s surrounding rock by using the multipoint displacement meter in the construction, the corresponding mechanical porameters of the surrounding rock are gotten through the network simulation.