Signalized intersections have a very important effect on control and management of urban traffic. Delay, otherwise, is one of the most significant parameters which reflects the performance of an intersection. With the inclusion of advanced driving behaviors and capabilities of driverless vehicles into our lives, it is predicted that traffic delays and other problems will decrease and traffic performance of signalized intersections and arterials will increase. In this study, the behavioral characteristics of human drivers (reaction time, acceleration and deceleration etc.) are obtained from the data collected by field studies and the characteristics of autonomous vehicles are determined based on assumptions defined in literature. In the study, it is also assumed that autonomous vehicles have fully automation (level 5 that can communicate with the system and the environment).
Traffic volume, delay and queue length data are collected from the field at evening peak hours from İzmir Narlıdere Dokuz Eylül Conservatory intersection in İzmir. Observed data is used for calibration of Vissim microsimulation program. Several scenarios are prepared in order to reflect the effect of autonomous vehicles, with penetration rates of 0%, 50% and 100% of autonomous vehicles only in the southern approach of the intersection. The results of the scenarios are compared and the impacts of autonomous vehicles on signalized intersections are examined. The results are evaluated in terms of delay and queue lengths. Results showed that, autonomous vehicles may increase delay at signalized intersection if they are not connected with surrounding environment.
Anahtar Kelimeler: Signalized Intersections, Autonomous Vehicles, Delay