Ant Colony Optimization Algorithm for Signal Coordination of Oversaturated Traffic Networks
Emily Zechman, Luca Quadrifoglio, Rahul Putha, Texas A&M University, May 2010, 58 pp. (169113-1)
Traffic congestion is a daily and growing problem of the modern era in mostly all major cities in the world. Increasing traffic demand strains the existing transportation system, leading to oversaturated network conditions, especially at peak hours. Oversaturation occurs when queues of vehicles fill the streets approaching intersections and interfere with the performance of adjacent upstream intersections. Traffic conditions, measured based on the overall throughput of vehicles and total travel time, can be improved by an effective employment of intelligent transportation system techniques. While a significant amount of research has been devoted to the development of signal control algorithms under normal traffic conditions, a relatively small number of studies have explicitly considered oversaturated conditions. The overall objective of this study was to investigate the effectiveness of the Ant Colony Optimization (ACO) algorithm in solving the traffic signal control problem under oversaturated conditions. Due to its ability to reach optimality conditions and identify acceptable solutions efficiently, ACO was a good candidate for a practical use. This research compared the performance of ACO to that of another heuristic method, the genetic algorithm (GA). The methods were applied to identify signal control strategies for two example networks. The results demonstrate that ACO was able to identify fit solutions more reliably than the GA-based approach.
Keywords: Transportation Planning, Traffic Signal Timing, Ant Colony Optimization, Genetic Algorithm
ENTIRE REPORT (Adobe Acrobat File – 747 KB)