中国机械工程 ›› 2026, Vol. 37 ›› Issue (6): 1508-1517.DOI: 10.3969/j.issn.1004-132X.2026.06.023

• 工程前沿 • 上一篇    下一篇

基于多目标约束的狭小空间便携式机器人轨迹优化方法

刘金锋1(), 顾世民1, 张占虎2, 李苏1, 陈宇1, 王学敏3, 钱天龙3   

  1. 1.江苏科技大学机械工程学院, 镇江, 212000
    2.中船动力(集团)有限公司, 镇江, 212002
    3.扬州中远海运重工有限公司, 扬州, 225200
  • 收稿日期:2025-06-06 出版日期:2026-06-25 发布日期:2026-07-17
  • 通讯作者: 刘金锋
  • 作者简介:刘金锋*(通信作者),男,1987年生,教授、博士研究生导师。研究方向为船舶智能制造关键使能技术。E-mail: liujinfeng@just.edu.cn
  • 基金资助:
    国家自然科学基金(52371324)

Trajectory Optimization Method for Portable Robots in Confined Spaces Based on Multi-objective Constraints

LIU Jinfeng1(), GU Shimin1, ZHANG Zhanhu2, LI Su1, CHEN Yu1, WANG Xuemin3, QIAN Tianlong3   

  1. 1.School of Mechanical Engineering,Jiangsu University of Science and Technology,Zhenjiang,Jiangsu,212000
    2.CSSC Power(Group)Co. ,Ltd. ,Zhenjiang,Jiangsu,212002
    3.COSCO Shipping Heavy Industry (Yangzhou) Co. ,Ltd. ,Yangzhou,Jiangsu,225200
  • Received:2025-06-06 Online:2026-06-25 Published:2026-07-17
  • Contact: LIU Jinfeng

摘要:

针对船舶狭小空间焊接时作业空间受限、焊枪姿态约束多、轨迹干涉风险高、焊接可达性差等挑战,以机械臂最短时间完成焊接为目标,提出基于改进遗传粒子群算法(IGA-PSO)的时间优化方案以优化焊接轨迹。构建便携式机器人、工件和场景的三维模型,明确机器人运动逻辑,建立焊接轨迹模型并确定焊接工艺;综合考虑焊接时间和可达性,设计多目标约束适应度函数,并建立时间优化与可达率目标函数;结合遗传和粒子群算法,对惯性权重引入线性递减和指数递减机制,对学习因子设计探索、开发和收敛阶段,对变异进行非线性调整,以提高算法的性能。通过算法测试、仿真验证和现场验证对该方法进行了案例验证,结果表明,优化后的机械臂位移、速度、加速度曲线平滑、无突变,且焊接可达率达到90%,验证了IGA-PSO算法的有效性。

关键词: 多目标优化, 狭小空间焊接, 路径优化, 改进遗传粒子群算法, 自适应机制

Abstract:

Aiming at the challenges in ship narrow-space welding, such as the restricted workspace, multiple constraints on torch posture, high risk of trajectory interference, and poor welding accessibility, a time-optimization scheme was proposed for welding trajectory optimization based on an improved IGA-PSO, to achieve welding with the robotic arm in the shortest time. Three-dimensional models of the portable robots, workpiece, and working scene were constructed, the motion logic of the robots was clarified, and the welding trajectory model and processes were established. A multi-objective constrained fitness function was designed by comprehensively considering both welding time and accessibility, and the objective functions for time optimization and accessibility rate were formulated. By integrating the genetic and particle swarm algorithms, improvements were introduced: a linearly decreasing and exponentially decreasing mechanism for the inertia weight was adopted; the learning factors were designed for exploration, exploitation, and convergence stages; and the mutation operation was adjusted nonlinearly, thereby enhancing the algorithm's performance. Algorithm testing, simulation verification, and on-site validation were carried out through a case study to verify the proposed method. The results show that the optimized robotic arms exhibit smooth displacement, velocity, and acceleration curves without abrupt changes, and the welding accessibility reaches 90%, which verifies the effectiveness of the IGA-PSO.

Key words: multi-objective optimization, narrow-space welding, path optimization, improved genetic algorithm-particle swarm optimization(IGA-PSO), adaptive mechanism

中图分类号: