基于遗传算法和神经网络的隧道围岩位移智能反分析  

Intelligent back-analysis of tunnel surrounding rock displacement based on genetic algorithm and neural network

作　　者：黄戡[1,2] 刘宝琛[1] 彭建国[2] 冯德山[3] 丁国华[2] 王跃飞[2] 

机构地区：[1]中南大学土木建筑学院,湖南长沙410075 [2]湖南省交通规划勘察设计院,湖南长沙410008 [3]中南大学地球科学与信息物理学院,湖南长沙410075

出　　处：《中南大学学报（自然科学版）》2011年第1期213-219,共7页Journal of Central South University:Science and Technology

基　　金：西部交通建设科技项目(20033179802)

摘　　要：基于正交试验设计和FLAC3D建立的学习样本以及测试样本,通过工程现场获取的围岩位移信息,用神经网络建立待反演参数与围岩位移之间潜在的映射关系。研究结果表明:利用该神经网络的仿真预测功能,结合遗传算法搜索反演参数的最优解,从而实现位移反分析;可将反演结果反馈于隧道支护结构的设计,实现隧道的信息化施工与设计。The learning samples and test samples were built based on orthogonal experimental design and FLAC3D numerical simulation, using the project site for the surrounding rock displacement information. The potential mapping between parameters and surrounding rock displacement was established using neural network. The results show that the optimal solution of inversion parameters can be derived to achieve the displacement back analysis combined with the genetic algorithm and prediction function of the nerve network. Then inversion results can be the feedback for the design of tunnel support structure to achieve the tunnel construction and design of information technology.

关 键 词：位移反分析 遗传算法 神经网络 正交试验 

分 类 号：U459.2[建筑科学—桥梁与隧道工程]