Self-organization phenomenon and the edge of chaos in traffic flow dynamics

Abstract

According to the complexity theory, a self-organization phenomenon might exist when the phase shifts from order (non-congested) to chaos (congested) state within a system. In this paper, cellular automaton (CA) traffic simulator is applied to investigate the self-organization forming process as the traffic density grows. Real traffic data are used to examine whether the power law, an indicator of self-organization phenomenon, exists or not. Our empirical evidence on a two-lane freeway shows that self-organization phenomenon appears when the density exceeds 45% (or 60 pcu/km); however, the edge of chaos is at a density level of 33.8% (or 45 pcu/km), which should be the starting control point (e.g., ramp metering) for recurrent congestion.