摘要
行车时间估计和最优路径选择是智能交通系统中的研究热点,特别是对于车辆导航系统更具有深远的意义.首先以传统的交通流理论为基础,采用间接模型和动力学模型进行行车时间估计,通过仿真实验比较了两模型的优劣,并使用实测数据分析得到的车流量信息对动力学模型进行改进.然后使用Dijkstra算法寻找出静态状态下的最优路径,再结合前面建立的时间估计模型,给出了适用于动态随机状态下的路径寻优算法,用于解决路段行车时间期望随出发时刻动态变化的问题.最后指出了交通实时信息对解决动态随机最优路线问题的重要性,并结合卡尔曼滤波算法对路段相关的情况作了进一步讨论.
Travel time estimation and choosing optimal routing are the research hotspot in the intelligent traffic system （ITS）, especially they have the profound value to vehicle navigation system. The travel time is firstly estimated using indirect model and dynamics model based on the traditional traffic flow respectively, and the two models have been compared with emulation mode. After the road - test data have been analyzed, the tr＇~c flow information is extracted to improve on the dynamics model. Then the Dijkstra algorithm is used to find the static state optimal routing, and the algorithm for dynanrical state optimal routing is finding by using the improved dynamics model and Dijkstra algorithm, we use it to solve the problem about the ex- pectation of travel time dependent on departure time. Finally we find the real time traffic information is very important to solve the dynamic and stochastic optimal routing problem, we use the kalman filter to solve the correlated routing condition.
出处
《数学的实践与认识》
CSCD
北大核心
2006年第7期31-43,共13页
Mathematics in Practice and Theory
