Catastrophe theory and patterns in 30-second freeway traffic data— Implications for incident detection

This paper analyses 30-second data on flows, occupancies, and average speeds of freeway operations near bottlenecks caused by incidents. These data are seen to be very poorly explained by the conventional traffic flow theory, but to accord quite well with a recently proposed model based on catastrophe theory. One of the key areas in which the catastrophe theory model fits the data better than conventional models is with regard to the transitions to and from congested operation upstream of incidents. The data suggest that such transitions are characterized by a fairly gentle change in occupancy, and a fairly constant flow, but a sudden, sharp, change in speed. This relatively smooth transition in occupancy, which could not be investigated with earlier data, lends additional support to the catastrophe theory model. On the strength of these results, the outline of an alternative logic, based on the catastrophe theory model of the flow-occupancy-speed pattern, is suggested for freeway incident detection.