交通流分配中利用矩阵迭代法计算最短路径的一点改进  

One Improvement to Find the Nearest Route by Matrix Iteration Method in Traffic Flow Distribution

作　　者：郭瑞军[1] 王晚香[1] 

机构地区：[1]大连交通大学交通运输工程学院

出　　处：《大连交通大学学报》2008年第4期41-44,共4页Journal of Dalian Jiaotong University

摘　　要：全有全无的最短路径法是进行交通流分配的一种方法,矩阵迭代法可以用来计算加权有向图的最短路径,对于一个n节点的加权有向图,传统方法最多须迭代n-1次,而利用改进方法,即在迭代过程中直接由D2r′计算出D2r′+1,最多迭代次数只需└log2n-1┘或└log2n-1┘+1次,对于大规模节点数的加权有向图尤其适用.The method of all- or- nothing nearest route is one way to distribute traffic flow. The matrix iteration method can be used to find the nearest route in a figure with directions and weights. The largest number of iteration may be n - 1 by traditional method to the figure with n points. However,the largest number of iteration is ＇log2^n-1＇or＇log2^n-1＇＋1 by calculating D^2r＇＋1 form D^2r＇ . The improvement is especially evident to this kind of figure with a large number of points.

关 键 词：交通流分配 矩阵迭代法 距离矩阵 最短路 

分 类 号：U491[交通运输工程—交通运输规划与管理] O241.6[交通运输工程—道路与铁道工程]