基于行程时间多步预测的实时路径导航算法  

Real-time navigation algorithm based on multi-step travel time prediction

作　　者：李进燕[1] 朱征宇[1] 刘琳[1] 刘微[1] 

机构地区：[1]重庆大学计算机学院,重庆400044

出　　处：《计算机应用研究》2013年第2期346-349,353,共5页Application Research of Computers

基　　金：国家科技支撑计划资助项目(2007BAH08B04)

摘　　要：针对现有车辆导航算法仅考虑单一数据,使所得路径实际行程时间比预期更长的问题,首先建立了基于卡尔曼滤波理论的行程时间多步预测模型;其次,提出了综合利用实时数据、行程时间多步预测数据及历史数据的实时路径导航算法,并改进了其实现的核心算法Dijkstra_pred。实验结果表明,基于三类数据的实时路径导航算法所得路径的实际行程时间从整体上优于仅采用实时数据的导航算法,且路径变化较少。In view of the existing algorithms only considering one type of data and longer real travel time than prediction travel time, firstly, this paper proposed multi-step travel time prediction model based on Kalman. Secondly, it presented the route navigation algorithm which based on real-time data, multi-step prediction data and historica~ data. In the meantime, it de- signed the core algorithm Dijkstra_pred. Experiments show that the travel time calculated by the proposed route guidance algo- rithm is superior to the one by the algorithm based on real-time data, and route calculated based on the algorithm proposed in this paper changes less than the one considering only real-time data.

关 键 词：智能交通系统 动态路径规划 车辆实时导航 行程时间多步预测 卡尔曼滤波理论 DIJKSTRA算法 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]