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 ›› 2023, Vol. 31 ›› Issue (12): 203-214.doi: 10.16381/j.cnki.issn1003-207x.2022.0461

• • 上一篇    

冷链物流配送的绿色车辆路径模型及其求解算法

周鲜成1,2,蒋涛营2,贺彩虹3(),王莉1,吕阳1   

  1. 1.湖南工商大学智能工程与智能制造学院, 湖南 长沙 410205
    2.湖南工商大学前沿交叉学院, 湖南 长沙 410205
    3.湖南工商大学会计学院, 湖南 长沙 410205
  • 收稿日期:2022-03-07 修回日期:2022-06-30 出版日期:2023-12-15 发布日期:2024-01-06
  • 通讯作者: 贺彩虹 E-mail:hecaihong1991@163.com
  • 基金资助:
    国家自然科学基金资助项目(71972069);湖南省社会科学成果评审委员会项目(XSP20YBC024);湖南省高校物流系统优化与运作管理科技创新团队项目(湘教通[2019]379号)

Green Vehicle Routing Model and Its Solution Algorithm in Cold-chain Logistics Distribution

Xian-cheng ZHOU1,2,Tao-ying JIANG2,Cai-hong HE3(),Li WANG1,Yang LV1   

  1. 1.School of Intelligent Engineering and Intelligent Manufacturing,Hunan University of Technology and Business,Changsha 410205,China
    2.School of Advanced Interdisciplinary Studies,Hunan University of Technology and Business,Changsha 410205,China
    3.School of Accounting,Hunan University of Technology and Business,Changsha 410205,China
  • Received:2022-03-07 Revised:2022-06-30 Online:2023-12-15 Published:2024-01-06
  • Contact: Cai-hong HE E-mail:hecaihong1991@163.com

摘要:

针对冷链物流配送的绿色车辆路径问题,引入新鲜度衰减函数计算配送过程中冷链产品的货损;考虑载重、行驶速度、低温制冷等因素对油耗和碳排放率的影响,确定油耗和碳排放的度量函数;根据配送时间和顾客满意度的关系,建立顾客满意度函数。在此基础上,以车辆使用成本、货损成本、油耗和碳排放成本、惩罚成本之和最小化和顾客平均满意度最大化作为优化目标,构建双目标数学模型,设计了一种基于NSGA-II和变邻域搜索的混合算法进行求解。算例仿真结果表明,构建的模型和提出的算法能在多个优化目标之间取得平衡,降低车辆使用成本和货损,减少油耗和碳排放,提高顾客满意度。

关键词: 冷链物流配送, 绿色车辆路径问题, NSGA-II, 变邻域搜索

Abstract:

In recent years, fresh food e-commerce in China has developed rapidly, with a significantly growing demand for cold-chain logistics distribution. In order to promote energy conservation and emission reduction in cold-chain logistics, the Green Vehicle Routing Problem in Cold-chain Logistics Distribution (GVRPCLD) has attracted increasing attention from both academia and industry. The GVRPCLD discussed in this paper involves set constraints, including (1) distribution center locations, (2) capacity-constrained homogeneous refrigerated vehicles, (3) customers’ number, locations, demands, time windows, service times, and (4) a low temperature kept in the cabin during distribution. The optimal vehicle routing plan is expected to realize the dual goals of total cost minimization and customer satisfaction maximization. Specifically, there are five types of costs that are included in the total cost, which are vehicle usage cost, deterioration cost, cooling cost, fuel consumption and carbon emission cost, and piecewise penalty costs for earliness and tardiness of deliveries. In the design, a freshness decay function is firstly applied to calculate the deterioration cost during distribution. Based on the comprehensive mode emission model (CMEM), the measurement functions of fuel consumption and carbon emission are determined, with the consideration of the influence of vehicle load, speed, low temperature and other factors on fuel consumption and carbon emission rate. Then, a customer satisfaction function is established according to the relationship between delivery time and customer satisfaction. Finally, a dual-objective GVRPCLD optimization model is constructed. In order to solve the GVRPCLD model, a hybrid algorithm (VNSNSGA-II) is designed in this paper, which is based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II) and the Variable Neighborhood Search (VNS). The basic idea of the algorithm can be described as follows. Firstly, the VNS is used to optimize some chromosomes in the population to expand the solution space. Secondly, the NSGA-II is performed to obtain non-dominated sorting and crowding degree. Thirdly, based on the individual's rank and crowding distance, the top 50% of chromosomes are obtained and retained as new parent population. Finally, the Pareto-optimal set is obtained by iteration. The simulation results show the following. For one thing, the GVRPCLD model can achieve multi-objective optimization among logistics cost, deterioration cost, fuel consumption and carbon emissions, and customer satisfaction. Furthermore, it is found that logistics cost is negatively correlated with customer satisfaction, indicating that the delivery plan makers can choose different delivery options according to their preference. Last but not least, the proposed VNSNSGA-II is tested, and it is confirmed that the algorithm can contribute to a Pareto-optimal set.

Key words: cold-chain logistics distribution, green vehicle routing problem, NSGA-II, variable neighborhood search

中图分类号: