基于XGBoost算法的大直径穿黄隧道施工期管片上浮研究

陈健, 靳军伟, 李新潮, 杨公标, 李明宇, 靳倩倩

doi:10.3973/j.issn.2096-4498.2023.S1.009

隧道建设 >

2023 , Vol. 43 >Issue S1: 72 - 80

DOI: https://doi.org/10.3973/j.issn.2096-4498.2023.S1.009

研究与探索

基于XGBoost算法的大直径穿黄隧道施工期管片上浮研究

展开

(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. 中国海洋大学环境科学与工程学院，山东青岛 266100; 4. 郑州大学土木工程学院，河南郑州 450001； 5. 龙湖集团控股有限公司, 北京 100012)

陈健（1973—），男，山东泗水人，2021年毕业于中国海洋大学，能源与环保专业，博士，正高级工程师，主要从事地下工程与大盾构隧道施工技术及科研工作。Email: chenjian1018@163.com。*通信作者：靳军伟， Email: jinjunwei@zzu.edu.cn。

网络出版日期: 2023-08-24

收起

Segment Uplift of LargeDiameter Tunnel Crossing Yellow River During Construction Based on XGBoost Algorithm

Expand

(1. China Railway 14th Bureau Group Corporation Limited, Jinan 250101, Shandong, China; 2. China Railway Construction Underwater Tunnel Engineering Laboratory, Jinan 250101, Shandong, China;3. College of Environmental Science and Engineering, Ocean University of China, Qingdao 266100, Shandong, China;4. School of Civil Engineering, Zhengzhou University, Zhengzhou 450001, Henan, China; 5. Longfor Group Holdings Limited, Beijing 100012, China)

Online published: 2023-08-24

Fold

摘要

为解决大直径盾构隧道面临的施工期盾尾管片上浮问题。针对济南黄河隧道项目，提出了基于XGBoost算法的大直径泥水平衡盾构隧道施工期管片上浮计算框架。通过采用主成分分析法将地层参数降维，采用R-reliefF算法对管片上浮的影响因素进行特征提取及数据预处理工作，从而建立用于管片上浮分析的数据集。进而使用XGBoost算法对大直径隧道管片上浮进行计算，并与随机森林算法预测结果进行了对比。结果表明本文所采用的计算框架得到的结果能较好地反映隧道管片施工期的上浮特征，同时发现XGBoost算法对于管片上浮过程的预测效果比随机森林更好。研究成果对于大直径隧道施工过程中的管片变形预测及控制有较好的指导意义。

关键词： 大直径隧道; 盾构施工; 管片上浮; 机器学习; XGBoost算法; 预测分析

本文引用格式

陈健, 靳军伟, 李新潮, 杨公标, 李明宇, 靳倩倩 . 基于XGBoost算法的大直径穿黄隧道施工期管片上浮研究[J]. 隧道建设, 2023 , 43(S1) : 72 -80 . DOI: 10.3973/j.issn.2096-4498.2023.S1.009

Abstract

Shield tail segment uplift is a common phenomenon encountered in largediameter shield tunneling through rivers. As a result, a case study is conducted on a shield tunnel crossing the Yellow river, and a segment uplift calculation framework based on XGboost algorithm is proposed for largediameter tunnel bored by a slurry shield. The dimension of formation parameters is reduced by principal component analysis, and RreliefF algorithm is used to extract the features and preprocess the factors affecting segment uplift, so as to establish a data set for segment uplift analysis. Furthermore, the XGBoost algorithm is used to calculate the uplift of largediameter tunnel segments, and the results are compared with those of random forest algorithm. The results show that the calculation framework used in this study can better reflect the uplift characteristics of tunnel segments during construction, and the XGBoost algorithm has a better prediction effect than random forest for the uplift process of tunnel segments. The research results have a good guiding significance for the prediction and control of segment deformation in the construction process of largediameter tunnel.

Key words： largediameter tunnel; shield construction; segment uplift; machine learning; XGBoost algorithm; forecast analysis

Options

文章导航

〈

〉

模态框（Modal）标题

摘要

本文引用格式

Abstract