Large traffic jam formation induced by multiple crossings in city network

Abstract

Crossing effects have not been considered in conventional large-scale macroscopic transportation network models. A large-scale macroscopic network model for vehicular flow is presented to take into account the crossing effect. We consider a simple directed network in which multiple y-directional roads cross a single x-directional road. Crossing y-directional roads interact with x-directional road through crossings. When a bottleneck or a blockage is set in city traffic network, traffic jams are formed and propagate in the transportation network through crossings. We show how jams propagate through crossings in the directed network. Jams are localized at a low mean density, extend over the network with increasing mean density, propagate towards the whole network at a high mean density, and major jams occur by chain reaction. It is shown how the extension of jams over the network changes with the mean density and the position of the bottleneck. The fundamental diagrams (flow-density relationship) are derived. Successive jamming transitions by chain reaction occur when the blockage is set.