Operations Research and Management Science ›› 2018, Vol. 27 ›› Issue (4): 72-82.DOI: 10.12005/orms.2018.0087

• Theory Analysis and Methodology Study • Previous Articles     Next Articles

The Cash-in-transit Routing Optimization Problem with Multi-type Risks

XU Guo-xun, LI Yan-feng, LI Jun, YANG fang   

  1. School of Economics and Management, Southwest Jiaotong University, Chengdu 610031, China
  • Received:2016-06-06 Online:2018-04-25

现金押运路线问题——基于多类型的风险

徐国勋, 李妍峰, 李军   

  1. 西南交通大学 经济管理学院,四川 成都 610031
  • 通讯作者: 李妍峰(1980-),女,副教授,博士生导师,研究方向:物流优化,交通优化。
  • 作者简介:徐国勋(1984-),男,博士生,研究方向:物流优化;李军(1967-),女,教授,博士生导师,研究方向:物流与供应链。
  • 基金资助:
    国家自然科学基金项目(71571150,71361006);教育部人文社会科学研究项目(14YJA630026);四川省哲学社会科学重点研究基地项目(QGXH15-05);中央高校基本科研业务费项目(26815WCX03);成都市科技局项目(2015-RK00-00038-ZF)

Abstract: Based on cash logistics industry, this paper introduces a routing optimization problem with multi-type risks. The problem is formulated as a mixed integer linear programming problem to balance transportation cost, transit risk and risk of cash in hand. A hybrid genetic algorithm adopted in this paper is a hybrid meta-heuristic combining the exploration capabilities of GA with efficient local search based improvement heuristics and diversity management mechanism. This method consists of a GA method to determine routing decision and a greedy method to determine the risk decision. The numerical studies are adopted to analyze problem characteristics and the method performance. The results show that the hybrid genetic algorithm can solve much larger problems, produce high quality solutions and strike a satisfying balance between the computing time and the solution quality.Multi-type risks link with the vehicle routing. Customer expect demand links with risk of cash in hand.

Key words: multi-type risks, vehicle routing problem, hybrid genetic algorithm, greedy method

摘要: 以人民币现金押运为研究背景,考虑了一种基于多类型风险的现金押运路线问题,以在途风险成本、库存现金风险成本以及运输成本为优化目标,建立了混合整数线性规划模型,并提出了一种基于多样化策略和改进邻域搜索的混合遗传算法,其中遗传算法对押运路线进行选择,贪心算法用来求解各类风险指标。数值实验分别对问题特性和算法性能进行了分析。实验结果表明:1)混合遗传算法能求解更大规模的问题,得到较好的解,并很好地平衡了运行时间和求解质量;2)多类型风险影响了行驶路线;3)客户的期望需求影响了库存现金风险。

关键词: 多类型风险, 车辆路径问题, 混合遗传算法, 贪心算法

CLC Number: