转向约束网络中的对偶最短路径树原理及其原型算法-交通运输工程学报.pdf

第8 卷　第4 期 交 通 运 输 工 程 学 报 Vol.8　No.4 2008 年8 月 Journal of T raffic and Transportation Engineering Aug.2008 :1671-1637(2008)04-0084-06 任　刚, 王　炜 ( , 　210096) 　:为比较有无转向约束条件下最短 径特征及其搜索算法的异同点, 基于对偶图理论证明了 转向约束网络中从单 个源点到所有弧的最短 径集构成其对偶网络的生成树, 提出了对偶最短 径树(DSPT)概念, 并利用其分析算法之间的关系。研究结果表明:转向约束下的现有求解方法包 括弧标号算法、节点标号算法和对偶网络法都可以统 一到DSPT 算法框架内, 而且与无转向约束 的最短 径树(SPT )算法在 径搜索策略上是相同的;对于转向约束网络中的最短 径问题可建 立一个DSPT 原型算法, 结合各种SPT 标号技术能设计出更多的有效算法。 :交通网络;对偶最短 径树;对偶图;转向约束;原型算法 :U491.13　　　:A Principles and its prototype algorithm for dual shortest path tree in networks with turning constraints REN Gang, WANG Wei (Jiangsu Provincial Key Laboratory of T ransportation Planning and M anagement, Southeast University, Nanjing 210096, Jiangsu, hina) Abstract:To analyze the difference and similarity between the shortest path characteristics and searching algorithms with and without turning constraints, dual graph theory was studied, the spanning tree of dual network was constructed with the shortest path set from a source node to all arcs in turning constraint network, the concept of dual shortest path tree (DSPT )was put forward and used for algorithm comparison.Analysis result shows that the present solution methods with turning constraints, including arc-labeling algorithm, node-labeling algorithm and dual-network method, can be unified into the same DSPT algorithmic frame, and have the same path searching strategies as the shortest path tree(SPT )algorithm without turning constraints;a prototype DSPT algorithm can be developed for the shortest path problem in networks with turning constraints, and more efficient algorithms can be designed according to different SPT labelin