Title:Traffic Response Functions: Patterns, Propagation and Congestion

Abstract:We further pursue a recently tested approach to the study of transport systems. Using empirical data collected on German motorways, we investigate the use of response functions as a tool to study traffic dynamics in motorway networks. We use a response function that measures the change in traffic observed at a given location over time conditioned on congestion at another location. By introducing several velocity-dependent congestion indicators, we are able to work out the basic characteristics of the flow and density responses to congestion. Our results demonstrate that these responses capture the propagation of congestion, the acceleration phases of traffic participants during congestion, and the correlation between the three fundamental traffic observables: flow, density, and velocity. Furthermore, a direct comparison with the underlying data reveals the potential for characterizing traffic patterns with response functions. A relationship between the shape of the response functions and the traffic oscillation during a traffic jam can be demonstrated. Moreover, our findings allow us to determine the propagation velocity of congestion from responses. Importantly, in the course of our analysis, we also identify a scaling property that results in the collapse of the response functions onto a single curve.