Chain reaction of traffic breakdowns in coupled-cycle networks

Abstract

We present a dynamic model for major traffic jams in urban-scale networks. We investigate the traffic breakdown in the directed network consisted of cycle graphs when an intersection is closed to vehicles. We show how the traffic breakdown propagates from cycle graph to cycle graph. The deadlock (jam) of vehicles is induced by the closed intersection, a local breakdown occurs at an early stage, and the breakdown propagates toward the whole network. The macroscopic dynamic equations of vehicular densities are derived by using the speed-matching model. The densities and currents (flows) on all roads are obtained numerically. We show that the chain reaction of traffic breakdowns occurs when a density is higher than a critical value, while it does not propagate over the network at lower density than the critical value. The chain reaction of traffic breakdowns induces the major traffic jam. The jamming transition between free traffic and major traffic jam is found in the traffic network.