中国机械工程 ›› 2021, Vol. 32 ›› Issue (21): 2606-2616.DOI: 10.3969/j.issn.1004-132X.2021.21.010

• 智能制造 • 上一篇    下一篇

预防维护下装配线平衡的多目标重启变邻域搜索算法

赵联鹏1,2;唐秋华1,2;张子凯1,2;蒙凯1,2   

  1. 1.武汉科技大学冶金装备及其控制教育部重点实验室,武汉,430081
    2.武汉科技大学机械传动与制造工程湖北省重点实验室,武汉,430081
  • 出版日期:2021-11-10 发布日期:2021-11-25
  • 通讯作者: 唐秋华(通信作者),女,1970年生,教授、博士研究生导师。研究方向为生产过程规划与调度、制造过程监测与控制、现代优化方法与算法。E-mail:tangqiuhua@wust.edu.cn。
  • 作者简介:赵联鹏,男,1997年生,硕士研究生。研究方向为智能算法与生产调度。E-mail:zhaolpie@163.com。
  • 基金资助:
    国家自然科学基金(51875421)

Multi-objective Restart Variable Neighborhood Search Algorithm for Assembly Line Balancing Considering Preventive Maintenance

ZHAO Lianpeng1,2;TANG Qiuhua1,2;ZHANG Zikai1,2;MENG kai1,2   

  1. 1.Key Laboratory of Metallurgical Equipment and Control,Ministry of Education,Wuhan University of Science and Technology,Wuhan,430081
    2.Hubei Key Laboratory of Mechanical Transmission and Manufacturing Engineering,Wuhan University of Science and Technology,Wuhan,430081
  • Online:2021-11-10 Published:2021-11-25

摘要: 针对预防维护下的装配线平衡问题,提出了一种带有重启策略的多目标变邻域搜索算法,以优化正常工作、设备维护情形下的节拍与工序调整。算法结合启发式与随机方法得到较优初始解;设计并筛选出寻优能力较强且具有互补性的四类邻域算子及其搜索策略,以更好地进行全局探索与局部开发。为促进Pareto前沿推进,提出了一种具有自适应能力的重启算子,以便根据问题规模调整重启代数阈值、参考寻优进程扩大搜索空间。该算法机制简单且无固定参数,实验结果表明该算法能够获得具有竞争性的非支配解集。

关键词: 装配线平衡, 预防维护, 变邻域搜索, Pareto优化, 自适应重启

Abstract: Aiming at the assembly line balancing problems considering preventive maintenance, a multi-objective variable neighborhood search algorithm with restart strategy was proposed to optimize task adjustments and cycle times under normal work and equipment maintenance scenarios. The proposed algorithm combined heuristics and stochastic methods to obtain a better initial solution. Four types of neighborhood operators with strong optimization ability as well as complementarity and their search strategy were designed, selected and effectively combined for better global exploration and local exploitation. In order to promote the advancement of Pareto front, a restart operator with adaptive capability was proposed to adjust the iteration threshold according to the scale of the problems and to expand the search space based on the optimization processes. The proposed algorithm is simple and has no fixed parameters. Experimental results show that the algorithm may obtain a set of non-dominated solutions with competing performance.

Key words: assembly line balancing, preventive maintenance(PM), variable neighborhood search(VNS), Pareto optimization, self-adaptive restart

中图分类号: