Simulation of the Impact of Connected and Automated Vehicles at a Signalized Intersection

Date of Award


Degree Name

M.S. in Civil Engineering


Department of Civil & Environmental Engineering & Engineering Mechanics


Advisor: Deogratias Eustace


Intersections are locations with higher likelihood of crash occurences and sources of traffic congestion as they act as bottlenecks compared with other parts of the roadway networks. Consequently, connected and automated vehicles (CAVs) can help to improve the efficiency of the roadways by reducing traffic congestion and traffic delays. Since CAVs are expected to take control from drivers (human control) in making many important decisions, thus they are expected to minimize driver (human) errors in driving tasks. Therefore, CAVs potential benefits of eliminating driver error include an increase in safety (crash reduction), smooth vehicle flow to reduce emissions, and reduce congestion in all roadway networks. Since CAV implementations are currently in early stages, researchers have found that the use of traffic modeling and simulation can assist decision makers by quantifying the impact of increasing levels of CAVs, helping to identify the effect this will have on future transportation facilities. The main objective of the current study was to simulate the potential impacts CAVs may have on traffic flow and delay at a typical urban signalized intersection. Essentially, to use a microscopic traffic simulation software to test future CAV technology within a virtual environment, by testing different levels of CAVs with their associated behaviors across several scenarios simulated. This study tested and simulated the impact of CAVs compared with conventional vehicles at a signalized intersection. Specifically, I analyzed and compared the operations of the signalized intersection when there are only conventional vehicles, conventional vehicles mixed with CAVs, and when there are only CAVs. The most current PTV Vissim 11 software was used for simulating different percentages of three different types of CAVs and conventional vehicles in the traffic stream at the intersection. These are three different levels of automated vehicles that are already installed in PTV Vissim 11, which are AV cautious, AV normal, and AV all-knowing. All these automated vehicles were tested in different scenarios in this study. Real data from an existing signalized intersection in the city of Dayton, Ohio were used in the PTV Vissim software simulation. The traffic count data used in the Vissim intersection model were for morning peak hour. The existing signal timing data for the intersection used were first optimized using Synchro. The results from Vissim simulation show that CAVs could reduce the queue delay by about 12%, the stopped delay by about 17%, the vehicle travel time by about 17%, and the queue length by about 22%. Because of that, CAVs can substantially reduce congestion at urban signalized intersections.


Civil Engineering, Transportation, Engineering, Connected and Automated Vehicles, Autonomous Vehicles, CoEXist, Traffic Simulation, Signalized Intersection, Intersection Delay

Rights Statement

Copyright © 2019, author