Name of Faculty Sponsor
Asma Azizi
Faculty Sponsor Email
aazizi@kennesaw.edu
Publication Date
2024
Abstract
Traffic congestion is an enduring problem for major metropolitan areas, such as Atlanta, GA. Our goal is to understand the nature of traffic congestion patterns in the highway system of Cobb County in Atlanta, GA. We created a road network representative of the Cobb County highway system and then superimposed a degree-based SIR model to simulate traffic congestion on that network. The model’s parameters, propagation and dissipation rates, were estimated using empirical traffic data, which are vehicles’ speed time series and speed limit of each road in the network. We then conducted a local sensitivity analysis of the model’s key parameters, confirming each parameter’s modest impact on the cumulative number of congested roads and the number of congested roads at peak time. Then, we used optimal control theory to identify the most effective control function in reducing traffic jams in Cobb County. Our findings showed that low levels of control did lower total congestion but mimicked the uncontrolled congestion behavior. However, increasing the level of control dropped congestion and changed its behavior by diminishing the peak. That is, the congestion time series monotonically decreased to zero. This paper provides additional support for the fact that traffic behavior is accurately predicted by SIR modeling and suggests that there exists a precise level of traffic control that eliminates traffic congestion propagation behaviors, given enough resources.
