基于模糊神经网络的公路隧道洞口段施工阶段风险评估  

Evaluation of Highway Tunnel Portal Construction Risk Based on the Fuzzy Neural Network Method

作　　者：黄越[1] 龚珍[1] 邓祥辉[1] 李鹏[1,2] 王睿 HUANG Yue;GONG Zhen;DENG Xiang-hui;LI Peng;WANG Rui(School of Civil and Architecture Engineering,Xi＇an Technological University,Xi＇an 710032,China;Shunde District of Foshan City Country Garden Property Development Co.Ltd.,Foshan 528000,China)

机构地区：[1]西安工业大学建筑工程学院,陕西西安710032 [2]佛山市顺德区碧桂园物业发展有限公司,广东佛山528000

出　　处：《工程管理学报》2018年第3期119-123,共5页Journal of Engineering Management

摘　　要：公路隧道洞口段埋深较浅,围岩风化严重,加上降雨的影响,很容易发生边仰坡失稳、洞口坍塌等事故,因此研究隧道洞口段施工风险对于保证施工安全极为重要。为了降低公路隧道洞口段施工阶段风险,分析了影响公路隧道洞口段施工阶段的风险及风险因素,运用层次分析法构建了公路隧道洞口段施工阶段风险评估指标体系,利用模糊层次综合评判法结合相关资料建立了样本数据,并对样本数据进行主成分分析;将模糊数学理论、主成分分析法与神经网络相结合建立了风险评估模型,提出了隧道洞口段施工阶段风险评估方法。选取45座公路隧道施工实例作为样本进行训练分析,以另外5座隧道作为评估目标进行评估,并与模糊层次综合评判法所得结果比较,两种方法评估结果误差较小。With the shallow buried depth of high tunnel portal, the rock weathering, and the influence of rainfall, it is easy to occur accidents such as the slope instability and the cave collapse at tunnel. Therefore, it is very important to study the construction risk of tunnel entrance section to ensure the safety of construction. To achieve it, this paper estimates the risk by analyzing influencing factors. The risk assessment index system of highway tunnel portal construction is built by using the Analytic Hierarchy Process. The sample data is established with the fuzzy hierarchy comprehensive evaluation method and related data. A risk assessment model is established by combining the fuzzy theory, the principal component analysis and neural network. The risk assessment method for the highway tunnel portal construction is verified by using 45 highway tunnels as samples. Furthermore, this study uses another 5 tunnels to validate the model. The result indicates that the error of results is smaller when compared with the fuzzy level comprehensive evaluation method.

关 键 词：公路隧道 洞口段 施工阶段 模糊神经网络 风险评估 

分 类 号：U455[建筑科学—桥梁与隧道工程]