中文摘要：

结合南京机场线地铁项目,进行了邻近隧道爆破开挖作用下既有隧道的爆破振动速度现场监测试验,应用BP神经网络,建立了既有隧道爆破振动速度的预测模型,并与多种经验公式预测进行了比较分析.结果表明：BP神经网络爆破振动速度预测数据与试验监测数据拟合较好,相比于经验公式预测,具有误差小、精度高的特点.研究成果可为邻近隧道爆破开挖作用下既有隧道爆破振动控制和新建隧道爆破开挖方案完善提供理论参考.

英文摘要：

Based on Nanjing airport line metro project, field monitoring tests were carried out on existing tunnel vibration response induced by blasting excavation of adjacent tunnel. A prediction model of blasting vibration velocity of existing tunnel was established using BP neural network, which was compared with a variety of empirical formulas. The results indicated that the prediction data of blasting vibration velocity by BP neural network were in a good coincidence with experimental data. Compared with empirical formula, BP neural network had characteristics of fewer error and high accuracy. The research could provide a theoretical reference for controlling blasting vibration of existing tunnel induced by blasting excavation of adjacent tunnel and improve blasting excavation program of new tunnel.