优化多式联运问题的一种广义最短路方法研究

Research on a Generalized Shortest Path Method of Optimizing Intermodal Transportation Problems

多式联运问题是目前引起普遍关注的一个研究课题,但针对多式联运网络的性质及其相关算法的研究,尤其是寻求最佳运输路线方法的研究,各类相关文献仍涉及很少。本文首先回顾多式联运问题的理论研究现状,并分析了Reddy(1995)构建的总运输成本最小化条件下的多式联运模型。在此基础上,提出一种求解最佳运输路线的广义最短路法,即通过构建多式联运网络多重图,将运输过程中的数据、信息和图中的节点、边关联起来,然后对运输费用和中转费用进行分析估计,并通过在联运网络图中加入虚拟的发、到站,使得该问题可用Dijkstra算法进行求解,从而获得广义费用最少的联运方案。最后,将该方法和后动态规划法同时应用于由5个城市及3种运输方式构成的多式联运算例求解,通过实际对比分析,证实该优化方法的计算复杂度不高于后动态规划法,从而验证了该优化方法的有效性。

Intermodal transportation is a research topic of great interest at present. However, little work has been done to investigate the properties and related algorithms of the intermodal network,in particular the method of finding an optimal transport path. This paper reviews the current research situation,and analyzes the intermodal transportation model with the minimized total transport costs established by Reddy（1995）. Basing on this,this paper proposes a generalized shortest path method. That is, by constructing an intermodal-network multi-ply graph which interrelates the data and information in transit with the nodes and sides of the graph, adding virtual origin and destination stations to the graph, and then analyzing transport costs and transfer costs, this paper applies the Dijkstra algorithm to solve the intermodal optimal transport path problem so as to get the intermodal transport plan with the least generalized costs. At last, this method and the backward dynamic programming method are both used to settle a numeral example which is composed of five cities and three transport modes. The comparative results prove that this method is no more than the backward dynamic programming method in calculation complexity, which verifies the effectiveness of this optimal method.