城市地下空间的高强度开发,使得地下管线、隧道及其它地下建、构筑物日趋密集。随着盾构施工穿越已建隧道的概率不断增大,盾构近距离或超近距离穿越已建构筑物的保护问题已然不可避免。本文从地表变形控制角度,通过BP神经网络技术建立地表下沉、地质信息、隧道和盾构机的几何参数与施工参数之间非线性关系的研究,借助变形控制指标优化施工参数,从而有效指导穿越段工程的顺利进行。 The intensive development of underground space in cities makes underground pipelines, tunnels and other underground structures and structures increasingly dense. With the increasing probability of shield tunneling through tunnels already constructed, the protection of the shields passing through the constructed structures at close or super-close distances is inevitable. In this paper, the nonlinear relationship between the geometric parameters of surface subsidence, geological information, tunnel and shield machine and construction parameters is established by BP neural network technology from the perspective of surface deformation control, and the construction parameters are optimized by means of deformation control index, so as to effectively guide Crossing the smooth progress of the project.