Abstract:
Traffic congestion at signalized intersections has been a recurring problem,
especially at urban arterials during peak hours. Many signalized intersections are
rapidly being converted into unconventional median U-turn intersections in Pakistan.
The Unconventional Median U-turn (UMUT) is one of the popular alternatives used to
mitigate congestion at signalized intersections. This research is an attempt to assess this
low-cost "indirect right turn" technique. The performance of a 4 -legged signalized
intersection and an unconventional median U-turn design have been compared by
creating several scenarios based on different traffic demands, percentages of right-turn
traffic, and median U-turn offsets. The signal timing of signalized intersections was
optimized using the Synchro software, while the traffic simulation software VISSIM
was used in the study to simulate the intersections. Python was used as a programming
language via API COM interface to generate different scenarios of conventional and
unconventional median U-turns in VISSIM. The evaluation parameters that were taken
into consideration for the comparison of both types of intersections included average
travel time delay, fuel consumption, and emissions. In MATLAB, vehicular delay
equations and travel time models were developed for all movements in an UMUT
design. These models were created based on simulation traffic data obtained from
VISSIM. The approach utilized Generalized Linear Models (GLMs) with exponential
relationships to accurately capture the associations between different input variables
and output parameters. Internal validation procedures were carried out to ensure the
reliability of the models.
