The proper evaluation of transportation improvements requires the estimation of daily traffic flow profiles. This research investigated the temporal variations of daily traffic flow profiles on urban access-controlled roadways. A daily traffic flow profile modeling procedure was developed.;Hourly traffic flow data and study site characteristics were obtained for 42 locations in Texas. This information was gathered for as many years as possible between 1969 and 1995 and daily traffic flow profiles were calculated. The study site characteristics of interest included: annual average daily traffic (AADT) per lane, county population, roadway orientation, and location.;Daily traffic flow profiles on urban access-controlled roadways typically have four intervals. The consistent definition of these intervals allows their attributes to be modeled. Twelve straight-line approximations of each daily traffic flow profile were completed. The best-fit approximation approach defined peak period attributes with the hours that had volumes within 25 percent of their most appropriate peak-hour volume. Transitions between the peak and off-peak periods were one or two hours wide.;The defining attributes of the straight-line approximations were modeled. AADT per lane was the focus of these models. Peak-period durations were modeled with respect to AADT per lane, and the morning peak-period start hour and evening peak-period end hour were based on peak-period duration. Interval magnitudes were related to interval duration or the database average. A seven-step procedure was developed that combined the model results for an estimation of daily traffic flow profiles.;The validity of the profile attribute models was tested. It was found that the majority of the models estimated their straight-line approximation counterparts relatively well. Most of the average interval durations and magnitudes were not significantly different than those originally approximated.;The results of the proposed approach were compared to those of the current profile approximation methodology and the calculated daily traffic flow profiles at 15 validation study sites. The current approach approximated daily traffic flow profiles more accurately than the proposed approach. The gross nature of the straight-line process within the proposed approach overwhelms the advantages it offers and the estimation abilities of the individual profile attribute models developed.
展开▼
