Modeling Driver Behavior At Signalized Intersections With Red Light Camera

A motorists' decision whether to pass or stop at the intersection during the yellow signal interval depends on a number of factors, such as speed, geometric characteristics, driver's attitude, to name a few. The decision with respect to the yellow signal can result in red light running or rear end and side collisions. Motorists' behavior at red light camera intersections during exposure with a yellow signal may be seen as a binary decision in which case the two main decisions are either to come to a stop or cross the intersection.

In this dissertation, a discrete choice model of the stopping probability is developed using vehicles' actual speed and location from the stop line when the motorist is exposed to the yellow signal. A binary choice model is developed using the probability of stopping to the yellow signal as a function of actual approach speed, distance from intersection, and presence of camera. The existence of the dilemma zone is estimated using dilemma zone plots developed from the probability of stopping vs. distance from stop line during the yellow interval.

The dissertation also presents a new approach to calculate the change in capacity resulting from drivers stopping at the intersection during the yellow interval. Using field data from Baltimore, Maryland it is shown that the capacity of camera equipped intersections may be lower than that at intersections without cameras.