关键词: 智能交通, 路径诱导, 短时交通流预测, 广域诱导模型, 时空边界条件, Logit 模型, 动态流量分配

Abstract:

In order to improve the efficiency of vehicle traffic, based on the predictive guidance strategy, this paper puts forward the queue length as the constraint condition of traffic guidance, uses the short- term traffic volume of wavelet neural network to predict the road section where the blocking event occurs, and then uses the time-space boundary conditions of wide area guidance to classify the nodes and define the length of the guidance period of the event road section, and then establishes the wide area guidance. Then, the event area road network is divided into districts to further determine the location of the guidance starting point of the model, and the Logit path selection model based on the path scale is introduced as the guidance path selection method. Finally, the load balance of the road network is realized by the iterative flow distribution method. Through the example, it is proved that the guidance method can effectively alleviate the road traffic congestion and improve the traffic efficiency of the road network.

Key words: intelligent traffic, route guidance, short-term traffic flow prediction, wide-area induced model, spatiotemporal boundary conditions, Logit model, dynamic traffic distribution