考虑多源信息不确定性的TBM利用率预测  

Prediction of TBM utilization considering multi-source information uncertainty

作　　者：李庆民 闫长斌[3] 姚西桐 杨公标 杨风威 杨继华 LI Qingmin;YAN Changbin;YAO Xitong;YANG Gongbiao;YANG Fengwei;YANG Jihua(China Railway 14th Bureau Group Corporation Limited,Jinan 250101,China;China Railway Construction Underwater Tunnel Engineering Laboratory,Jinan 250101,China;School of Civil Engineering,Zhengzhou University,Zhengzhou 450001,China;Guangxi Communications Design Group Co.Ltd.,Nanning 530029,China;Key Laboratory of Water Management and Water Security for Yellow River Basin,Ministry of Water Resources(under construction),Zhengzhou 450003,China)

机构地区：[1]中铁十四局集团有限公司,山东济南250101 [2]中国铁建水下隧道工程实验室,山东济南250101 [3]郑州大学土木工程学院,河南郑州450001 [4]广西交通设计集团有限公司,广西南宁530029 [5]水利部黄河流域水治理与水安全重点实验室(筹),河南郑州450003

出　　处：《中南大学学报（自然科学版）》2023年第10期4043-4056,共14页Journal of Central South University:Science and Technology

基　　金：国家自然科学基金资助项目(41972270);中铁十四局集团有限公司科技研发计划课题(9137000016305598912021C03);水利部黄河流域水治理与水安全重点实验室(筹)研究基金资助项目(2022-SYSJJ-06)。

摘　　要：隧道掘进机(TBM)施工对地质条件非常敏感,一旦发生事故会造成严重的工期延误和巨大的经济损失。本文在对TBM利用率影响因素进行统计分析和对输入参数进行优选的基础上,以兰州水源地建设工程输水隧洞双护盾TBM施工实例为依托,将天牛须搜索优化方法与增强回归树算法进行耦合,提出一种考虑多源信息不确定性的TBM利用率预测模型(BAS-BRT);将该模型预测结果与粒子群优化-增强回归树耦合模型(PSO-BRT)预测结果进行对比分析,并通过现场实测数据验证TBM利用率预测模型BAS-BRT的有效性及其对典型地质段施工风险的适应性。研究结果表明:天牛须优化算法的自适应特点能够真实反映TBM施工中出现的不确定性问题,增强回归树算法可实现模型的全局最优化迭代,TBM利用率预测模型BAS-BRT具有较高的预测精度、较好的泛化性能,同时具有良好的并行处理能力与鲁棒性。TBM construction is very sensitive to geological conditions,and once an accident occurs,it will cause serious delays in the construction period and huge economic losses.On the basis of the statistical analysis of factors affecting the utilization of TBM and the optimization of input parameters,and based on the measured data of double-shield TBM in the water conveyance tunnel of the Lanzhou water source construction project,a BAS-BRT prediction model of TBM utilization considering the uncertainty of multi-source information was proposed by coupling the beetle search optimization method with the boosted regression tree algorithm.And the TBM utilization prediction results of BAS-BRT model were compared with those of particle swarm optimization−boosted regression tree coupled model(PSO-BRT).The validity of the BAS-BRT prediction model of TBM utilization rate and its adaptability to the construction risk of typical geological sections were verified by the field measured data.The results show that the self-adaptive characteristics of the beetle optimization algorithm can truly reflect the problem of uncertainty in TBM construction.The boosted regression tree algorithm can realize the global optimization iteration of the model.The BAS-BRT prediction model of TBM utilization has high prediction accuracy,good generalization performance and it has good parallel processing ability and robustness as well.

关 键 词：隧道掘进机(TBM) 设备利用率 不确定性 天牛须搜索方法 增强回归树 施工风险 

分 类 号：U455[建筑科学—桥梁与隧道工程] TV554[交通运输工程—道路与铁道工程]