为了更加真实的反映岩土体整体特性主要力学参数的取值,以长春地铁某暗挖区间现场监测位移值为依据,通过敏感性分析,确定了该地区对地表变形影响最大的力学参数为弹性模量E和内摩擦角φ,运用Madis GTS三维正算模型、BP神经网络模型组成的智能位移反分析系统对其进行了反分析。运用反分析得到的参数值进行数值模拟,将得到的监测断面位移值与实测值进行了对比,并根据断面的位移云图对隧道围岩的稳定性进行了评价,验证了反分析结果的合理性和准确性。
In order to determine the value of the main mechanical parameters that could truely reflect the global property of rock and soil ,the mining section of Changchun subway was taken as the engineering background ,and it was confirmed that the mechanical parameters with the greatest impact on the surface deformation were the elasticity modulus E and in-ternal friction angle φ,based on the sensitivity analysis of the field monitoring displacement data .And then the intelli-gent displacement back analysis system formed by the Midas GTS 3D forward solution model and BP neural network was applied for the back analysis of these parameters .The new parameters obtained from the back analysis were used in the simulation analysis ,and the displacement value was calculated and then compared with the monitoring data .According to the displacement cloud picture of the section ,the stability of the tunnel surrounding rock was evaluated ,and the rational-ity and accuracy of the back analysis result was verified .