巴郎山隧道围岩稳定性前期分析

Ba-Lang Hill Tunnel Surrounding Rock Stability Analysis in the Prior Period

以巴郎山隧道为例,在施工过程中对隧道围岩变化情况进行监控量测,本文采用BP神经网络模型和灰色理论GM(1,1)模型对围岩的前期变化进行分析。结果表明两种模型拟合预测精度均较高,能较好地反映隧道在施工过程中,水平位移收敛的变化情况,其中灰色模型拟合平均精度较优,并且没有误差较大的跳跃点,而神经网络模型拟合平均精度稍弱且局部误差较大,通过对比有利于施工者在开挖前期,选择适应性更强的模型对隧道稳定性进行判断。

In this paper, relying on the baron mountain tunnel, carries on the monitoring measurement in the construc-tion process to the changes of surrounding rock. This paper uses the BP neural network model and grey theory model to analyze changes in the early period of the surrounding rock. Results show that the two models fitting accuracy are high, can well re-flect the tunnel in construction process, the change of horizontal displacement convergence, the grey model has a better fitting average precision, and there is no jump point of error, but the neural network model fitting average precision is less and the lo-cal error is bigger, by comparing the builders in the early stage of the excavation and choose more flexible model of tunnel stabil-ity judgment.