Capacity Improvements Based on Dynamic Signal Timing at Signalized Road Intersections Using Genetic Algorithm

Abstract

Due to lack of public transportation, economic growth and the rapid increase in the number of populations, the number of vehicles in Erbil city has recently increased dramatically. This increment caused congestion on the roads and posed most of the loads on the traffic intersections, which already have limited capacity. As the intersection volume increases, the degree of saturation (X) and delay increases and, as a consequence, the Level-of Service (LOS) decreases, leading to instability and platoon formation at the intersection approaches. One of the reasons of congestion at intersections is due to the fixed signal timing during the day, while the traffic volume at the intersections varies from one hour to another. This research suggests solutions to vehicle delays at intersections through using a Genetic Algorithm (GA) based optimization model to adjust the signal parameters to fit the real time traffic condition. The suggested model tries to optimize the cycle length besides the green time per phase for different time lags throughout the day. The results showed distinct and remarkable improvements in the intersection capacity with 2 to 13 % along with an average delay of 28- 64 %.