Exploring urban traffic flow dynamics

Important statistical characteristics of traffic can be examined via the reconstruction of a so-called phase-space where each point in that space represents a state of traffic and vice versa. Statistical tools such as recurrence analysis can then be used to visualize the temporal evolution of traffic and to assess its non-stationarity structure as an indicator for irregular behaviour. In this paper we investigate, examine and analyze the short-term non-stationary features present in traffic flows based on the historical data available from stop-line detectors collected from a small number of intersections in the Melbourne CBD, Australia and demonstrate that recurrence analysis has considerable potential in identifying changes within the temporal behaviour of traffic.