DYNAMIC CLUSTERING AND PROPAGATION OF CONGESTION IN LARGE TRAFFIC NETWORKS

The problem of clustering in urban traffic networks has been mainly studied in static framework by only considering traffic conditions at a given time. Nevertheless, it is important to underline here that traffic is a strongly time-variant process and it needs to be studied in the spatiotemporal dimension. Investigating the clustering problem over time helps us understand and reveal the hidden information during the process of congestion formation and dissolution. The primary motivation of the paper is to study the spatiotemporal relation of congested links, observing congestion propagation from a macroscopic perspective, and finally identifying critical congestion regimes to aid the design of peripheral control strategies and improve mobility. The proposed approach utilizes a set of distinct and robust homogeneous components in the network called snakes, which represent a sequence of connected links with similar level of congestion. Firstly, we reduce the search space by selecting a sub-set of distinct snakes which cover different parts of the network. Secondly, a quadratic binary optimization framework is designed to find major skeleton of clusters from obtained distinct snakes by minimizing a heterogeneity index. Thirdly, we present how space and time correlations across clusters are captured with an iterative procedure that identifies the links with the highest degree of heterogeneity due to time dependent conditions and re-cluster them to guarantee connectivity and minimize heterogeneity.