带时间窗的多式联运作业整合优化算法  

Optimization algorithm of multimodal transportation with time window and job integration of multi-agent

作　　者：熊桂武[1,2] 王勇[1] 

机构地区：[1]重庆大学经济与工商管理学院,重庆400044 [2]四川外语学院国际商学院,重庆400031

出　　处：《系统工程学报》2011年第3期379-386,共8页Journal of Systems Engineering

基　　金：重庆市软科学资助项目(CSTC2010CE0089);教育部人文社会科学资助项目(10YJC630299)

摘　　要：快速、低成本的货物运输已成为现代物流的发展趋势,多式联运及多作业整合是物流企业减少运输时间和成本的有效方法.为此,首先建立了基于图状结构的带时间窗的多式联运及多代理人作业整合优化模型;然后提出了代理商选择和路径及运输方式协同优化的两层优化算法:第一层采用转移、交换和环形移动三种移动策略实现作业的分配,第二层采用基于正交试验的混合田口遗传算法实现路径及运输方式的组合优化;最后用具有不同时间约束的算例来考察模型和优化算法的有效性.结果表明,受时间约束的影响,作业整合只能在部分作业之间进行,并且随着时间窗约束的增强,物流运输总费用将增加,与实际情况基本相符,说明了模型及优化算法的可行性和有效性.The multimodal transportation and multi-job integration（MTMJI） are effective in reducing the transportation time and cost.Therefore,an optimaization model based on the graph structure is firstly formulated for the MTMJI with the time window.An optimized algorithm with two layers is then proposed after characterizing the formulated model.Three move strategies including the shift move,the swap move and the loop move are applied to effectively assign jobs to agents at the first layer.A hybrid genetic algorithm based on orthogonal experiment is adopted to find an optimal solution to the selection of route and transportation manner at the second layer.A numerical simulation with different time windows is finally performed to validate the formulated model and the proposed algorithm.The simulation shows that the integration of jobs can only be carried out among partial jobs due to the constraint effect of time window;the total cost will increase with the restraint intensity of time window,which is similar to the practical case.Furthermore,the results demonstrate the formulated model and proposed algorithm are effective and feasible.

关 键 词：第四方物流 时间窗 多代理人 作业整合 混合田口遗传算法 邻域搜索算法 

分 类 号：F572[经济管理—产业经济]