摘要
The accurate and timely traffic state prediction has become increasingly important for the traffic participants,especially for the traffic managements. In this paper,the traffic state is described by Micro-LOS,and a direct prediction method is introduced. The development of the proposed method is based on Maximum Entropy (ME) models trained for multiple modes. In the Multimode Maximum Entropy (MME) framework,the different features like temporal and spatial features of traffic systems,regional traffic state are integrated simultaneously,and the different state behaviors based on 14 traffic modes defined by average speed according to the date-time division are also dealt with. The experiments based on the real data in Beijing expressway prove that the MME models outperforms the already existing model in both effectiveness and robustness.
The accurate and timely traffic state prediction has become increasingly important for the traffic participants,especially for the traffic managements. In this paper,the traffic state is described by Micro-LOS,and a direct prediction method is introduced. The development of the proposed method is based on Maximum Entropy (ME) models trained for multiple modes. In the Multimode Maximum Entropy (MME) framework,the different features like temporal and spatial features of traffic systems,regional traffic state are integrated simultaneously,and the different state behaviors based on 14 traffic modes defined by average speed according to the date-time division are also dealt with. The experiments based on the real data in Beijing expressway prove that the MME models outperforms the already existing model in both effectiveness and robustness.
作者
SUN XiaoLiang1,2, JIA LiMin1, DONG HongHui1, QIN Yong1 & GUO Min3 1State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing 100044, China
2School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China
3Beijing Traffic Management Bureau, Beijing 100044, China
基金
supported by Beijing Science Foundation Plan Project(Grant No.D07020601400707)
the National High Technology Re-search and Development Program of China(Grant NO.2006AA11Z231)
