主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2026, Vol. 34 ›› Issue (11): 210-218.doi: 10.16381/j.cnki.issn1003-207x.2023.1833cstr: 32146.14.j.cnki.issn1003-207x.2023.1833

• • 上一篇    下一篇

基于同步循环取货的多行程汽车零部件入厂物流优化研究

陈剑(), 薛强, 周鸿, 马文静   

  1. 南京航空航天大学经济与管理学院,江苏 南京 211106
  • 收稿日期:2023-11-03 修回日期:2024-05-01 出版日期:2026-11-25 发布日期:2026-11-10
  • 通讯作者: 陈剑 E-mail:jchen@nuaa.edu.cn
  • 基金资助:
    国家自然科学基金项目(52075259);国家自然科学基金项目(72571139);中国博士后科学基金项目(2019M661839);中国博士后科学基金项目(2021T140320);江苏高校“青蓝工程”项目

Multi-trip Automotive Parts Inbound Logistics Optimization Based on Synchronous Milk-run Pickup

Jian Chen(), Qiang Xue, Hong Zhou, Wenjing Ma   

  1. College of Economics and Management,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China
  • Received:2023-11-03 Revised:2024-05-01 Online:2026-11-25 Published:2026-11-10
  • Contact: Jian Chen E-mail:jchen@nuaa.edu.cn

摘要:

为了保障汽车制造商有序生产及降低零部件入厂物流成本,本文提出了同步循环取货模式,综合考虑多行程路径优化、需求按零部件拆分、零部件同步交付等实际情况,构建了一个基于多行程路径优化的混合整数规划模型。针对所提的模型,设计了一个改进的自适应大规模邻域搜索算法,采用前向插入启发式方式生成初始解,设计了问题相关的破坏和修复算子,提出了一个基于扰动的后局部搜索,增强了算法的全局勘探和局部寻优能力。基于新能源汽车制造商数据生成数值算例,开展了算法性能实验和敏感性实验,结果表明,所提方法的良好性能,能够有效降低汽车零部件入厂物流成本。

关键词: 零部件入厂物流, 同步循环取货, 自适应大规模邻域搜索算法, 多行程路径优化

Abstract:

Parts inbound logistics refers to the process of managing the transportation, storage, and handling of parts or components as they move from suppliers to a manufacturer. This is particularly critical in the new energy vehicle industry, where the variety of models and personalized components requires optimized logistics for incoming parts to enhance competitive advantage. Synchronous pickup is one of the important means of parts inbound logistics for just-in-time production. Personalized demand for high and large parts needs to be supplied to mixed-model assembly lines on time. In order to ensure the parts timely delivery, direct delivery mode is usually used to ensure synchronous pickup. Although it can ensure the timeliness of transportation, it suffers from low logistical efficiency and high empty return rates.In order to achieve better synchronous pickup, a new synchronous milk-run mode is proposed by combining milk-run with synchronous pickup. Its specific process is given as follows. First, automotive manufacturer splits parts demand of synchronous workstations in a mixed-model assembly line and sends the demand to suppliers in advance. Second, the third-party logistics company adopts the multi-trip synchronous milk-run mode to pick up parts and deliver to synchronous workstations directly within strict delivery deadline.A mixed-integer programming model based on multi-trip vehicle routing problem is built by taking into account the characteristics of demand splitting of parts and synchronous delivery of parts. For the proposed model, an improved adaptive large neighborhood search (IALNS) algorithm is designed. It adopts a pushforward insertion heuristic to generate the initial solution and designs several problem-related destroy and repair operators to generate high-quality offspring solutions. Additionally, a new perturbation-based post-local search is proposed to improve global and local search capabilities. Numerical experimental results demonstrate excellent performance and speed of the proposed IALNS compared to Gurobi, variable neighborhood search and genetic algorithms in different-scale instances. Sensitivity experiments reveal that as delivery lead times become more lenient, the average vehicle travel distance initially increases before stabilizing gradually, while transportation costs show a decreasing trend. The proposed approach can effectively improve vehicle utilization efficiency, reduce inbound logistics costs for parts, and provide theoretical support for inbound logistics in automotive manufacturers.

Key words: parts inbound logistics, synchronous milk-run pickup, adaptive large neighborhood search, multi-trip vehicle routing problem

中图分类号: