Characteristics of traffic flow on urban expressway - A case study of Beijing

In this paper, the characteristics of traffic flow on urban expressway in Beijing are investigated. Based on the real data, some classical traffic flow models are used to explore the relationships among traffic flow, speed and time occupancy. The FCM (fuzzy C-means method) clustering algorithm is applied to classify the traffic flow state. The results show that the Greenshields model is the best suitable model for the mathematical description of traffic flow, i.e. traffic speed and time occupancy is in linear relation while traffic volume and time occupancy (traffic speed) is in parabolic relation. Four different traffic flow states, including free flow, steady flow, synchronized flow and congested flow, are identified according to the FCM clustering method. For the free traffic flow, the time occupancy cluster center is equal to 0.16, and the traffic volume cluster center is equal to 980 pcu/h. For the steady traffic flow, the time occupancy cluster center is equal to 0.30, and the traffic volume cluster center is equal to 1737 pcu/h. For the synchronized traffic flow, the time occupancy cluster center is equal to 0.50, and the traffic volume cluster center is equal to 1896 pcu/h. For the congested traffic flow, the time occupancy cluster center is equal to 0.79, and the traffic volume cluster center is equal to 1184 pcu/h. The analysis of traffic flow characteristics in this study can provide the method and data foundation to evaluate the traffic service level and enhance the traffic organization and management.