摘要
动态路径诱导系统(dynamic route guidance system,DRGS)是通过提供基于实时交通信息的最优路径来引导交通流的,因此,最优路径的求解是关键。而遗传算法具有全局寻优和潜在并行的特点,对求解最优路径具有一定优势。但采用序号编码方式进行遗传操作时会产生大量无效路径。文中结合城市道路交叉口左转、右转、直行等转向行为,设计了一种新的基于转向行为的编码方式,减少了染色体在交叉、变异时的无效路径的生成。算例表明,这种编码方式可以有效提高算法收敛性,更容易获得最优解。
The dynamic route guidance system （DRGS）guides the traffic flow by providing drivers with optimal route based on real-time traffic information, in order to alleviate traffic congestion and improve network efficiency. Optimal path algorithm is the key, and genetic algorithm has such characteristics as overall optimization and potential parallel. It has certain superiority in solving the optimal path problem. However, using the method of number encoding will provide massive invalid paths in heredity. Therefore, combined with the behaviors of turning left, turning right, and going straight in the intersection of urban road, a new encoding method based on turning behaviors was designed in this paper. It could reduce the invalid paths produced by overlapping and variation of chromosome. The algorithm shows that this encoding method can effectively prove the astringency of the algorithm, and is easier to obtain the optimal solution.
出处
《交通与计算机》
2007年第5期89-92,共4页
Computer and Communications
关键词
遗传算法
最优路径
动态路径诱导系统
染色体编码
genetic algorithm
optimal route
dynamic route guidance system
chromosome coding
