Postprint version. Published in Transportation Research Part B: Methodological, Volume 33, Issue 1, February 1, 1999, pages 25-42.
NOTE: At the time of publication, the author Robert Bertini was not yet affiliated with Cal Poly.
The definitive version is available at https://doi.org/10.1016/S0191-2615(98)00023-X.
Observations from two freeway bottlenecks in and near Toronto, Canada indicate that the average rate vehicles discharge from a queue can be 10% lower than the flow measured prior to the queue's formation. Absent any influences from downstream, the queue discharge flows exhibited nearly stationary patterns that alternated between higher and lower rates. These alternating flow patterns were especially evident at one of the two sites, although the feature occurred at both sites during periods that immediately followed the onset of upstream queuing; i.e. a queue's formation was always accompanied by a relatively low discharge rate followed later by a temporary surge in the discharge flow. When plotted cumulatively over time, however, the counts of discharging vehicles generally did not deviate by more than about 50 vehicles from a trend line of constant slope. Thus, the discharge flows are described as being `nearly' constant; i.e. they varied (slightly) about a fixed rate. At each site, this average discharge rate exhibited little deviation from day to day. The present findings came by visually comparing transformed curves of cumulative vehicle arrival number vs time and cumulative occupancy vs time measured at neighboring loop detectors. This treatment of the data provided clear presentations of some important traffic features and this facilitated a detailed study of bottleneck flows.
Civil and Environmental Engineering