Abstract
Considerable congestion may develop on urban and suburban freeways during peak day periods, especially when traffic demand is high. Traffic breakdown is identified when there is a steep rise in density levels and a simultaneous rapid drop in speeds during the phase transition from dense congested stable (DCS) flow to breakdown flow. The purpose of this study was to generate breakdown thresholds by modeling the process of freeway flow breakdown. This paper uses the mathematical property of log periodic oscillations (LPO) to model the traffic density over time. The LPO process has been suggested in the past for modeling stock market crashes and the occurrence of large earthquakes, but this is the first time it has been proposed for freeway breakdown modeling. The calibration results obtained threshold values of densities and corresponding speeds, which trigger the breakdown process. The critical point has been identified as the maximum density rate of change during the first cycle that follows the DCS flow regime. The resulting density thresholds are comparatively higher than the equal density line thresholds proposed by the Highway Capacity Manual (HCM) between levels of service E and F. The use of Athol's conversion factor between occupancy and density has a major impact on this finding, and further examination has shown a good match between the LPO and HCM density thresholds.
Get full access to this article
View all access options for this article.