基于人工神经网络的炭质岩隧道锚杆应力计算研究  

Bolt Stress Calculation of Carbonaceous Rock Tunnel Based on Artificial Neural Network

作　　者：王才进 骆俊晖 张涛[1] 段隆臣[1] 马冲[3] WANG Caijin;LUO Junhui;ZHANG Tao;DUAN Longchen;MA Chong(Faculty of Engineering China University of Geosciences(Wuhan))Wuhan 430074,China;Guangxi Communications Design Group Co.,Ltd.,Nanning 530029,China;School of Mathematics and Physics,China University of Geosciences(Wuhan),Wuhan 430074,Cina)

机构地区：[1]中国地质大学(武汉)工程学院,湖北武汉430074 [2]广西交通设计集团有限公司,广西南宁530029 [3]中国地质大学(武汉)数学与物理学院,湖北武汉430074

出　　处：《安全与环境工程》2020年第2期155-161,共7页Safety and Environmental Engineering

基　　金：国家自然科学基金青年基金项目(41807260);中央高校基本科研业务费专项资金项目(CUG170636、UCG170807);中国广西科技计划项目(桂科AC16380119、桂科AD19110124)。

摘　　要：锚杆在软弱围岩隧道加固中的应用非常广泛,为研究其受力特性,以广西某炭质岩隧道为依托,采用现场试验的方法对隧道监测断面锚杆应力进行了测试,并利用人工神经网络(ANN)分析方法,建立了基于ANN的隧道锚杆应力预测模型,通过将模型预测结果与实测结果进行比较,以验证所建模型的准确性和有效性。结果表明:所建模型以围岩应力、渗透压和围岩应变位移为输入参数,较为全面、合理地反映了锚杆应力的主要影响因素;模型计算所得预测值与实测值的相关系数R^2大于0.65,均方根误差RMSE低于0.65 MPa,方差比VAF大于80%,说明模型能够准确、有效地预测锚杆应力。该研究可为有效地估算复杂地质条件下隧道支护的锚杆应力提供参考。The bolt reinforcement has a wide application in the construction projects of tunnels with weak surrounding rocks.In order to study the stress characteristics of the bolts,this paper conducts a field test to monitor the changes of bolt stress in the tunnel section,which is surrounded by the carbonaceous rocks and located in Guangxi Province.Based on the Artificial Neural Network(ANN)analysis method,the paper develops a prediction model to estimate the bolt stress of the tunnel and also verifies the validity of the proposed model by comparing the predicted and measured values.The results show that the model can accurately and effectively predict the bolt stress.The model takes the stress,osmotic pressure and strain displacement of surrounding rock as input parameters,and reflects the main factors affecting the bolt stress comprehensively and reasonably.The correlation coefficient R^2 between predicted and measured values is greater than 0.65,the Root Mean Square Error(RMSE)is lower than 0.65 MPa,and the Variance Account For(VAF)is greater than 80%.The research can provide reference for effectively estimating the bolt stress of tunnel support under complex geological conditions.

关 键 词：炭质岩隧道 锚杆应力预测 现场测试 人工神经网络(ANN) 

分 类 号：X935[环境科学与工程—安全科学] TU43[建筑科学—岩土工程]