• 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) ›› 2025, Vol. 45 ›› Issue (12): 2298-2312.DOI: 10.3973/j.issn.2096-4498.2025.12.010

• 研究与探索 • 上一篇    下一篇

基于MIWOA-PSO算法的盾构管片拼装机轨迹规划

曾哲坤1, 胡明1, *, 汪珑2, 陈治进2, 周传璐1   

  1. (1.浙江理工大学机械工程学院, 浙江 杭州 310018; 2. 中铁十四局集团有限公司, 山东 济南 250101)
  • 出版日期:2025-12-20 发布日期:2025-12-20
  • 作者简介:曾哲坤(2000—),男,浙江杭州人,浙江理工大学机械工程专业在读硕士,研究方向为盾构管片拼装机技术。 E-mail: 17816074882@163.com。 *通信作者: 胡明, E-mail: huming@zstu.edu.cn。

Trajectory Planning for Segment Erector in Shield Tunneling Based on Multistrategy Improved Whale Optimization Algorithm and Particle Swarm Optimization Algorithm

ZENG Zhekun1, HU Ming1, *, WANG Long2, CHEN Zhijin2, ZHOU Chuanlu1   

  1. (1. School of Mechanical Engineering, Zhejiang Sci-Tech University, Hangzhou 310018, Zhejiang, China; 2. China Railway 14th Bureau Group Corporation Limited, Jinan 250101, Shandong, China)
  • Online:2025-12-20 Published:2025-12-20

摘要: 为提升盾构管片拼装机的工作效率,减小管片拼装过程中的运动冲击,提高管片拼装机的工作稳定性,提出一种基于多策略混合鲸鱼-粒子群规划算法(MIWOA-PSO)的时间-冲击最优轨迹规划方法。首先,对管片拼装机进行运动学分析,通过运动学逆解将笛卡尔空间的目标轨迹转换至关节空间。其次,采用5次NURBS(non-uniform rational B-splines)曲线对管片拼装路径进行插值,建立时间-冲击的双目标函数。为提升算法优化性能,对鲸鱼优化算法(WOA)进行多策略改进: 采用Tent混沌映射增强种群多样性,引入遗传算法的变异操作提升全局搜索能力,结合精英反向优化策略避免早熟收敛,并与粒子群算法(PSO)混合形成MIWOA-PSO算法。最后,通过5次NURBS生成初始平滑轨迹,再利用MIWOA-PSO算法优化时间参数,得到时间-冲击最优的运动曲线。结果表明: 1)在作业时间方面,优化后的轨迹总时长为23.23 s,相较于优化前的28 s缩短了4.77 s(17.04%),提高了管片拼装机的运行效率; 2)在运动平稳性方面,管片拼装机各关节的冲击均得到抑制,最大关节冲击值仅为极限冲击值的 33.58%,提升了管片拼装机的作业稳定性。

关键词: 盾构, 管片拼装机, 轨迹规划, MIWOA-PSO算法, NURBS曲线, 时间-冲击最优

Abstract: In this study, a time-jerk optimal trajectory optimization method based on a multistrategy improved whale optimization algorithm and particle swarm optimization (MIWOA-PSO) is proposed for enhancing the working efficiency of the segment erector in shield tunneling, reducing motion impact during segment assembly, and improving operational stability. Firstly, kinematic analysis of the segment erector is conducted, and the target trajectory in Cartesian space is mapped to the joint space via inverse kinematics solutions. Secondly, the segment assembly path is interpolated using quintic non-uniform rational B-spline (NURBS) curves, and an objective function with time and jerk as the optimization targets is established. Concurrently, the performance of the whale optimization algorithm (WOA) is optimized via multistrategy improvements: Tent chaotic mapping is employed to increase the population diversity, mutation operations from the genetic algorithm are introduced to enhance the global search capability, and an elite opposition-based learning strategy is incorporated to prevent premature convergence. The improved WOA is then hybridized with the PSO algorithm to form the MIWOA-PSO algorithm. Finally, an initial smooth trajectory is generated using quintic NURBS, and the time parameters are optimized via the MIWOA-PSO algorithm to derive a time-jerk optimal motion profile. The results indicate that: (1) in terms of operation time, the total duration of the optimized trajectory is 23.23 s, which is 4.77 s (17.04%) shorter than the pre-optimization time of 28 s, indicating improved operational efficiency of the segment erector; (2) regarding motion smoothness, the jerk in each joint is effectively suppressed, where the maximum joint jerk is only 33.58% of the limit, indicating enhanced operational stability of the segment erector.

Key words: shield, segment erector, trajectory planning, multistrategy improved whale optimization algorithm and particle swarm optimization algorithm, quintic non-uniform rational B-spline curve, time-jerk optimizatio