• CSCD核心中文核心科技核心
  • RCCSE(A+)公路运输高质量期刊T1
  • Ei CompendexScopusWJCI
  • EBSCOPж(AJ)JST
二维码

隧道建设(中英文) ›› 2026, Vol. 46 ›› Issue (5): 1109-1120.DOI: 10.3973/j.issn.2096-4498.2026.05.017

• 施工技术 • 上一篇    下一篇

长距离大直径复合地层盾构刀盘系统与掘进参数的多目标优化

王燚, 郑文宝, 罗刚, 刘建平   

  1. (中铁隧道集团二处有限公司,  河北 三河 065201)
  • 出版日期:2026-05-20 发布日期:2026-05-20
  • 作者简介:王燚(1994—),男,安徽池州人,2016年毕业于重庆交通大学,交通建设与装备专业,本科,工程师,现从事盾构施工技术管理工作。E-mail: 1214657406@qq.com。

Multi-Objective Optimization of a Large-Diameter Shield Cutterhead System and Tunneling Parameters in Complex Strata

WANG Yi, ZHENG Wenbao, LUO Gang, LIU Jianping   

  1. (The 2nd Engineering Co., Ltd. of China Railway Tunnel Group, Sanhe 065201, Hebei, China)
  • Online:2026-05-20 Published:2026-05-20

摘要:

为解决超长距离、超大直径、超高水压、超大埋深复合地层盾构隧道在局部夹砂卵石区段存在刀具磨损严重、掘进参数动态适配困难、施工效率与设备服役周期难以协同兼顾等问题,以上海—南京—合肥高速铁路崇太长江隧道工程为依托,提出刀盘系统优化设计—磨损状态监测—掘进参数智能协同控制一体化技术方案。在结构方面,构建“6主梁+6辅梁”高刚性刀盘骨架,配置由41把常压可更换撕裂刀和54把常压可更换切刀组成的立体式全轨迹刀具系统,并设置刀盘刀具磨损监测设施,以提升复杂地层条件下刀盘系统的抗磨损能力与可维护性。在控制方面,建立自适应多目标遗传算法(AMO-GA)模型,将总推力、刀盘转速、泥水压力、泥浆流量等关键参数纳入统一优化框架,并引入动态交叉、动态变异及地层风险指数(RFRI)权重调节机制,实现掘进效率与关键部件服役周期的协同优化。以0—130环为基线区段、131—250环为智能掘进应用区段进行对比分析。结果表明: 1)AMO-GA表现〖为宏观进尺提升,且在掘进过程中对质量控制良好,即参数调整响应时效由人工决策30 min以上缩短至90 s以内,转矩波动由±35%收敛至±12%,刀具百米磨损量由1.2 mm降至0.6 mm,密封件更换里程间隔由3 km 延长至8 km,泥浆循环利用率由65%提高至88%。2)刀具平均使用寿命提升超过50%,最高日进尺达到28 m,月平均进度约700 m,年维护成本降低62.5%。该技术体系能够有效增强复合地层盾构掘进过程中的参数响应能力、负载匹配能力和关键部件服役稳定性。

关键词: 复合地层, 盾构隧道, 刀盘, 掘进参数, 自适应多目标遗传算法

Abstract:

Super-large shield machines often encounter severe cutter wear, difficulty in the dynamic adjustment of tunneling parameters, and conflicts between tunneling efficiency and equipment lifespan when performing long-distance tunneling in composite strata with high water pressure and significant burial depth. To address these challenges, this study focuses on the Chongtai Yangtze River Tunnel of the Shanghai-Nanjing-Hefei High-speed Railway and proposes an integrated technical scheme that includes cutterhead system optimization, wear-state monitoring, and intelligent control of tunneling parameters. For structural design, a high-rigidity cutterhead framework featuring six main beams and six auxiliary beams is developed. A three-dimensional full-coverage cutter system is configured with 41 atmospheric-pressure replaceable ripping cutters and 54 atmospheric-pressure replaceable cutters. In addition, monitoring facilities for cutterhead and cutter wear are installed to enhance wear resistance and maintainability under complex geological conditions. In terms of control strategy, an adaptive multi-objective genetic algorithm (AMOGA) model is produced that incorporates key parameters, including total thrust, cutterhead rotation speed, slurry pressure, and slurry flow rate, into a unified optimization framework. Dynamic crossover, dynamic mutation, and a risk index-based weight adjustment mechanism are introduced to achieve coordinated optimization between tunneling efficiency and the lifespan of key components. Comparative analysis is performed using Rings 0-130 and 131-250 as the baseline and intelligent tunneling application sections, respectively. The results indicate the following: (1) AMO-GA significantly improves the advance rate and enhances control quality during tunneling. The response time for parameter adjustment is reduced from over 30 min under manual decision-making to within 90 s. The torque fluctuation range is reduced from ±35% to ±12%, the cutter wear per hundred meters decreases from 1.2 mm to 0.6 mm, the replacement mileage interval of sealing parts is extended from 3 km to 8 km, and the slurry recycling rate increases from 65% to 88%. (2) The average service life of cutters increases by over 50%, the maximum daily advance reaches 28 m, the average monthly advance is approximately 700 m, and annual maintenance costs are reduced by 62.5%. The proposed technical system effectively enhances parameter response capabilities, loadmatching performance, and the service stability of key components during shield tunneling in composite strata.

Key words: composite stratum, shield tunnel, cutterhead, tunneling parameters, adaptive multi-objective genetic algorithm