Empirical studies of unblocked index for urban freeway traffic flow states

People tend to know whether the traffic flow state in their desired route is smooth or the unblocked degree of traffic flow state in real life. To quantitatively measure the unblocked degree of traffic flow states for urban freeway, traffic flow states identification was firstly studied with fuzzy mathematics using speed data of urban freeway in Beijing, which were obtained from video capture technique. In this paper, fuzzy c-means clustering algorithm was used for developing membership functions. Traffic flow states were divided into four types, free flow, coherent moving flow, synchronized flow and jam. And then a computational method of unblocked index of traffic flow states was proposed from the viewpoint of similarity using max-min similarity statistical parameter and weighted summation. Free flow was chosen as the reference traffic flow state for the computational of similarity. Moreover, an entropy function was further developed for describing how the unblocked index change during traffic flow states transformation. Finally empirical analyses were discovered and studied based on field data.