基于ACO-SVM的TBM引水隧洞锚杆轴力预测研究  

Research on Axial Force Prediction of Bolt in TBM Headrace Tunnel Based on ACO-SVM

作　　者：王建 王利明 周振梁 路文甲 毛培良 Wang Jian;Wang Liming;Zhou Zhenliang;Lu Wenjia;Mao Peiliang(Xinjiang Erqis River Investment Development(Group)Co.,Ltd.,Urumqi 830000,Xinjiang,China;State Key Laboratory of Shield Machine and Boring Technology,Zhengzhou 450001,Henan,China;China Railway Tunnel Bureau Group Co.,Ltd.,Guangzhou 511458,Guangdong,China;Beijing Jiaotong University,Beijing 100000,China)

机构地区：[1]新疆额尔齐斯河投资开发(集团)有限公司,新疆乌鲁木齐830000 [2]盾构及掘进技术国家重点实验室,河南郑州450001 [3]中铁隧道局集团有限公司,广东广州511458 [4]北京交通大学,北京100000

出　　处：《岩土工程技术》2023年第5期532-537,共6页Geotechnical Engineering Technique

基　　金：国家重点研发计划(2020YFB2006804);河南省科技攻关(212102310270);新疆额尔齐斯河流域开发工程建设管理局科研课题(EQ076/FY057);中国中铁股份有限公司科技研究开发计划(2019-重大-10);中铁隧道局集团科技创新计划(隧研合2019-10)。

摘　　要：预测隧道锚杆轴力变化趋势,不仅能掌握隧道结构的安全状况,而且对隧道风险预警和应急响应至关重要。基于新疆YEGS输水工程喀双隧洞锚杆轴力监测数据,通过蚁群算法(ACO)和粒子群算法(PSO)优化支持向量机(SVM)模型,预测分析锚杆轴力的变化趋势,研究表明:相比PSO-SVM和传统的SVM预测模型,ACO-SVM预测模型在充分考虑隧道埋深、温度及作用时间等多项非线性影响因素后,求解的预测值与实测值更加接近,相对误差基本在15%以内,平均绝对百分误差仅为5.92,具有较好的鲁棒性,模型的稳定性和泛化能力更强,更加适合TBM隧道锚杆轴力变化趋势的预测分析,具有一定的工程应用和推广价值。Based on the monitoring data of anchor axial force in Ka-Shuang tunnel of Xinjiang YEGS water conveyance project,the change trend of anchor axial force were forecasted and analyzed through ant colony algorithm(ACO)and particle swarm optimization(PSO)to optimize the support vector machine(SVM)model.The research shows that ACO-SVM prediction model fully considers the tunnel buried depth compared with PSO-SVM and traditional SVM prediction model.After a number of nonlinear influencing factors such as temperature and action time,the predicted value is closer to the measured value,the relative error is basically within 15%,and the average absolute percentage error is only 5.92.The model has better robustness,stability and generalization ability,and is more suitable for the prediction and analysis of the variation trend of bolt axial force in TBM tunnel.It has certain engineering application and popularization value.

关 键 词：TBM 现场监测 锚杆轴力预测 支持向量机 蚁群算法 粒子群算法 

分 类 号：U45[建筑科学—桥梁与隧道工程]