隧洞开挖爆破振动监测与振速预测分析

Monitoring and Forecasting of Blasting Vibration Velocity in Tunnel Excavation

根据盛洪卿隧道泄水洞开挖爆破施工实际情况,选定合适的爆破监测方案。结合现场监测数据,分别采用经验公式线性回归法和BP神经网络法对爆破振动速度进行预测并将2种预测结果进行了比较分析。结果表明,线性回归法对地质条件具有依赖性,而BP神经网络方法可以较全面地考虑爆破振动速度的多种影响因素,且具有误差小、精度高等特性,因而应用BP神经网络方法预测爆破振动速度是有效可行的。

The suitable blasting monitoring program has been selected according to the actual condition of biasring excavation in Shenghongqing tunnel. Basing on the monitoring data,the blasting vibration velocity forecasting has been done by regression analysis and BP neural network;At the same time, the author had compared the results of two methods. The results indicate that the regression analysis method depends too much on geological conditions, while BP neural network method is able to consider the multi-factor of blasting vibration velocity, and it has the characteristics of a little error and high precision. Therefore, the BP neural network method is feasible in forecasting blasting vibration velocity.