Urban Road Traffic Bottleneck Identification

This paper proposes an identification method of traffic bottlenecks from aspects of static and dynamic bottleneck and predicts the dynamic bottleneck in the road network. First, traffic indicators were identified to find out static traffic bottleneck on urban road network, which were based on fuzzy reasoning mechanism. This paper used the road network around Weihai WEGO Square as a case study to realize this method. Then, the traffic parameters associated with dynamic traffic bottleneck were determined. According to that, association rules were mined to identify dynamic traffic bottleneck and were summarized. Finally, a BP-neural network model was established to predict dynamic bottle network using traffic data from a virtual road network. Results showed that this method could identify traffic bottlenecks of both types, and BP-neural network model could predict the traffic state of the next stage for road segments with 72.4% accuracy.