摘要
隧道围岩监控量测是隧道安全施工的重要保证,通过对隧道现场实测数据进行建模来分析围岩施工的变形特征,可以判断隧道围岩的稳定性,从而指导隧道的下一步施工。利用灰色系统理论GM(1,1)模型和BP神经网络模型对某隧道现场监控量测数据进行数据拟合与建模分析,并通过计算拟合残差选取更适合的预测模型来处理分析数据。由模型预测结果可知,BP神经网络模型所得到的拟合值更加贴合实际发展趋势,拟合精度更高,具有一定的适应性。
Monitoring and measuring for long tunnel surrounding rock is an important guarantee for the safety of tunnel construction.The data measured at tunnel construction sites can be used for modeling analysis of deformation characteristics during rock construction.The result of this analysis will be employed to determine the stability of tunnel surrounding rock,and provide a guidance for next phase of construction.This paper conducts data fitting and modeling analysis of the field monitoring and measuring data from a tunnel by using grey system theory GM（1,1）forecasting model and BP neural network principle.Through calculating fitting residuals,a more appropriate forecast model can be adopted to analyze the data of monitoring measurements.The forecasting result of the model reveals that fifting values absorbed from BP neural network algorithm are more adaptive to actual development trend and have better fifting precision as well as certain adaptability.
出处
《安全与环境工程》
CAS
2016年第3期152-157,共6页
Safety and Environmental Engineering
基金
国家自然科学基金项目(41374017)
关键词
长隧道
围岩变形
监控量测
灰色系统理论
BP神经网络
long tunnel
deformation of surrounding rock
monitoring measurement
grey system theory
BP neural network
