基于降雨量预测岩溶隧道施工掌子面涌水量的神经网络模型构建及软件开发  

Neural Network Model and Software Development on Prediction of the Water Inflow from the Working Face in the Karst Tunnel Construction Basing on Rainfall

作　　者：刘意立[1] 罗晋[1] 梁军平[1] 左静[2] 周亮 Liu Yili Luo Jin Liang Junping Zuo Jing Zhou Liang(School of Environment, Tsinghua University, Beijing 100084, China Lanzhou Jiaotong University, Lanzhou Gansu 730070, China Environmental Monitoring Station of Liangping, Chongqing 405200, China)

机构地区：[1]清华大学环境学院,北京100084 [2]兰州交通大学,甘肃兰州730070 [3]重庆市梁平县环境监测站,重庆405200

出　　处：《铁道建筑技术》2016年第12期33-36,共4页Railway Construction Technology

基　　金：国家自然科学基金(71173177);甘肃省青年科技基金项目(145RJYA242)

摘　　要：隧道施工掌子面涌水量的准确预测是保障施工安全的基础。本文利用神经网络模型,以预测目标日前6 d的降雨量数据和对应日期的掌子面涌水量作为神经网络模型的输入层参数,并根据未来降雨量的预测值对隧道施工掌子面涌水量进行预测。通过开发的可视化预测软件进行计算,该模型(软件)可对掌子面涌水量做出准确预测。In tunnel construction, the prediction of water inflow from the working face accurately is the basis to safety. The neural network model was adopted to forecast the water inflow in a designated day （6 d before the target date）, basing on the input parameters of correspondingly rainfall and inflow quantity of proximate days. The model was calculated by a visual software prepared by the writer and the result showed that the model （software） could make an accurate prediction of inflow quantity from the working face.

关 键 词：隧道 掌子面 涌水预测 神经网络模型 

分 类 号：U456.32[建筑科学—桥梁与隧道工程]