英语论文网

Intersections
At unsignalized intersections, three cameras mounted on tripods were used for data
collection. Two were Sony DV cameras that gave a time stamp in 1/30th of a second. This was
necessary since the gap acceptances and rejections required an accuracy of less than 1 second.
These two DV cameras (Cameras 1 and 3 in Figure 2) were placed at the ends of the
Figure 2. Typical Camera Coverage at Unsignalized Intersection
7
intersections looking at opposing and subject link flows. The third camera was placed to cover
the intersection area as sometimes the queue blocked the view of the other two cameras, thereby
preventing the recording of the gaps that were rejected or accepted. The DV cameras collected
the traffic volumes while the gap acceptance/rejections were collected from the center camera
and/or the DV cameras. Figure 2 shows the typical camera coverage that was followed for the
data collection at unsignalized intersections. The subject link and the opposing link are also
shown in Figure 2. Data were also collected for approximately 2 hours at each intersection
during either the morning or evening peak hours.
Data Reduction
The video data were processed to determine and/or verify information on arrivals, gap
acceptances, and traffic counts. Vehicle arrival times on both subject and opposing links were
recorded. In addition, vehicle turning movements were also recorded. While a left-turning
vehicle on the subject link was waiting for an acceptable gap, time gaps of opposing vehicle and
their acceptance or rejection were recorded.
Geometric and Traffic Data
The geometric and traffic characteristics of each intersection were identified and used as
input data for the event-based simulation program developed in this project. These data
included:
• number of approaches
• number of lanes on each approach
• turn attribute of each lane (left only, right only, etc.)
• volumes on each approach
• percentage of turns on each approach
• operating speed.
Critical Gap Data
The critical gap of the drivers plays an important role in the delays experienced by the
left-turning vehicles and therefore will determine whether the shared lane will be blocked or not.
This dictates the need for the separate left-turn lane. It was necessary to measure the critical gap
of the vehicles in the field to examine left-turn behavior.
As mentioned earlier, for both signalized and unsignalized intersections, the arrival time
at the stop line of each left-turning vehicle on the subject link was recorded. The arrival time at
the stop line of each opposing vehicle was also recorded. The difference in these times would be
the gap that is available for the left-turning vehicle. The left-turning vehicle would accept or
reject this available gap; therefore, all these gaps were measured and classified as “R” for
rejections and “A” for acceptances. A curve showing the rejections and acceptances was plotted
for that particular site. The intersection of the curve for A and R gave the gap above which
drivers would accept and below which drivers would reject, which is the critical gap at that
8
particular site. This was done for both signalized and unsignalized intersections. The procedure
described here can be found in Garber and Hoel (1999).
Assume that t1 and t2 are two t