Dynamic identification of urban traffic congestion warning communities in heterogeneous networks

摘要: Abstract Network-wide traffic control strategies (e.g. perimeter and route guidance) in urban networks have recently been mainly studied to relieve or postpone congestion based on the theory of macroscopic fundamental diagram (MFD). Nevertheless, these studies are mostly applied statically partitioned dynamic that fail fully consider state prediction, conflicting with strongly spatiotemporal variability objective active management (ATM). This paper proposes a methodology dynamically identify critical warning areas from heterogeneous road networks, which aids design efficient approaches. In methodology, directed weighted network is built base link connectivity real-time loads, travel time prediction method Kalman filter developed calibrate weight values undirected input for links. With information, community detection consists three consecutive steps developed. Firstly, it could capture emergence new definition seed intersection. Secondly, expansion regression each congested area be achieved objectives spatial compactness condition homogeneity. Thirdly, two-level merging algorithm adjacent different designed utilizing modularity model complex networks. The proposed validated using ground truth data downtown Jinan City China. results show algorithms can efficiently track evolutionary processes effectively detect test real network.