计算机工程与应用 ›› 2020, Vol. 56 ›› Issue (14): 240-249.DOI: 10.3778/j.issn.1002-8331.1904-0416

• 工程与应用 • 上一篇    下一篇

基于混合蚁群算法的异质车队低碳VRP研究

张明伟,李波,屈晓龙,郭盈   

  1. 1.天津大学仁爱学院 管理系,天津 301636
    2.天津大学 管理与经济学部,天津 300072
  • 出版日期:2020-07-15 发布日期:2020-07-14

Research on Low Carbon VRP of Heterogeneous Fleet Based on Hybrid Ant Colony Algorithm

ZHANG Mingwei, LI Bo, QU Xiaolong, GUO Ying   

  1. 1.Department of Management, Renai College of Tianjin University, Tianjin 301636, China
    2.School of Management & Economics, Tianjin University, Tianjin 300072, China
  • Online:2020-07-15 Published:2020-07-14

摘要:

针对货运车辆在配送调度过程中产生大量碳排放的问题,建立模型将多种影响碳排放量的因素协同优化。模型中考虑了不同载重量的异质车队,两个节点之间有多条道路的柔性路径,以及车辆重量随卸货而减少的动态负载等因素,以碳排放量、行驶时间和行驶路程为优化目标,并加入了节点需求时间窗、根据速度变化划分路段、交接和卸货时间的约束。提出了一种混合蚁群算法,利用蚁群算法信息素强度更新方式保持群体记忆性,利用粒子群算法的快速收敛特性增加计算效率。通过随机数值算例的仿真优化与对比分析,验证了算法和模型的有效性。

关键词: 车辆路径问题, 低碳, 异质车队, 柔性路径, 混合蚁群算法

Abstract:

In order to solve the problem of large amount of carbon emissions produced by freight vehicles in the process of distribution scheduling, a model is established to optimize the factors affecting carbon emissions synergistically. In the model, heterogeneous fleets of different rated loads, flexible path with multiple roads between two nodes, and dynamic loads that the vehicle weight decreases with unloading etc. are taken into consideration. Carbon emissions, travel time and travel distance are taken as optimization objectives, and the constraints of node demand time window, division of sections according to speed change, delivery and unloading time are added. A hybrid ant colony algorithm is proposed, which keeps the group memory by updating the pheromone intensity of ant colony algorithm, and improves the computational efficiency by using the fast convergence property of particle swarm algorithm. The validity of the algorithm and the model is verified by the simulation optimization and comparative analysis by random numerical examples.

Key words: vehicle routing problem, low-carbon, heterogeneous fleet, flexible path, hybrid ant colony algorithm