ART1神经网络在隧道围岩分类中的应用  

The application of ART1 neural network to the classification of surrounding rocks in tunnels

作　　者：李天斌[1] 王睿 

机构地区：[1]成都理工大学环境与土木工程学院 [2]中国民航机场建设总公司西南分公司,成都610202

出　　处：《成都理工大学学报（自然科学版）》2006年第5期455-459,共5页Journal of Chengdu University of Technology: Science & Technology Edition

基　　金：四川省杰出青年学科带头人培养计划资助项目(03ZQ026-045)

摘　　要：将自适应共振理论(ART)神经网络模型用于隧道围岩分类,改进了ART1神经网络的工作过程,通过自适应的学习记忆过程,建立了分类模型,有效地避免了人为主观因素的干扰。利用川藏公路二郎山隧道围岩分类样本对模型进行检验,结果表明,ART1神经网络模型性能良好,对隧道围岩分类的精度较高,是一种值得推广和应用的围岩智能分类方法。In this paper, the neural network model based on adaptive resonance theory （ART） is used for the classification of surrounding rocks in tunnels. The work process of ART1 neural network is ameliorated. The classification model for surrounding rock is built on it. As the classified samples, the data of surrounding rocks of Erlang Mountain tunnel on Sichuan-Tibet road are used to test the model. The result shows that ART1 neural network is good on function, and the results of surrounding rock classification are very accurate.

关 键 词：ART1神经网络 围岩分类 隧道 

分 类 号：U452.12[建筑科学—桥梁与隧道工程]