英语论文网

teps. The LTGAP is composed of three parts: (1) unsignalized
intersection simulation program, (2) signalized intersection simulation program, and (3) user
interface for data entry and display of results.
The unsignalized and signalized intersection simulation programs were written in C++,
and the interface was developed using Microsoft Visual Studio .Net program.
Validation of Simulation Model
Each event-based simulation model was validated using observed data from the field.
As mentioned earlier, the model for the unsignalized intersection was validated based on the
number of left-turn vehicle stopped, whereas the model for signalized intersections was validated
using maximum queue length and left-turn stopped delay.
Multiple simulation runs were made to account for variability in stochastic simulation
program output. A total of 100 runs were made for each test site, and the distribution of the
measures of effectiveness was compared with the field data using a histogram.
Development of New Guidelines
The guidelines developed by previous researchers were investigated, and the possibility
of developing guidelines based on pure operational savings (such as delay savings) was also
investigated by conducting a preliminary cost-benefit analysis. Separate guidelines were
developed for signalized and unsignalized intersections using the event-based simulation
program developed and validated in this study.
10
Guidelines for Unsignalized Intersections

The event-based simulation program was run for various combinations of input variables
to develop proposed left-turn lane guidelines. The follow levels of input variables were used to
generate new guidelines:
• percentage of left-turning vehicles = 3%, 5%, 10%, 20%, 30%
• operating speed = 40 mph, 50 mph, 60 mph
• advancing volume, VA, = 100 to 800 vehicles per hour (vph) by 50-vph increment
• opposing volume, VO, = 100 to 800 vph by 50-vph increment.
The maximum advancing and opposing volumes of 800 vph were used because 800 is the
practically highest acceptable volume that can be used for determining left-turn lanes at
unsignalized intersection. Traffic volumes of more than 800 vph would justify signalization of
the intersection. The criterion used for developing the guidelines was the percentage of time
through-vehicles were delayed. This was interpolated to find the volumes at their critical values.
This was plotted for all combinations of left-turn percentages. The critical percentages of time
that through vehicles were delayed were adopted from those used by Harmelink (1967).

Guidelines for Signalized Intersections
Pre-timed Signals
The program was run for all combinations of g/C (0.1 through 0.8 in 0.1 increments),
cycle length (60, 80, and 100 seconds), number of lanes (four and six lanes), and percentage of
left-turn vehicles (3%, 5%, 10%, 20%, and 30%). The capacity of the advancing volume, VA,
and the left-turn delay were calculated using the event-based simulation model. The proposed
guidelines recommended that a left-turn lane be installed when the volume-capacity (v/c) ratio
exceeded 0.85 or the delay exceeded 55 seconds/vehicle (LOS E cutoff). The guidelines were
developed in the form of tables.
Actuated Signals
The simulation program could be applied (using the LTGAP interface) to any actuat