Traffic signal optimization with Particle Swarm Optimization for signalized roundabouts

Abstract

At complex intersections, traffic congestion causes pollution and leads to accidents, high commute times and many other problems. Correct traffic signal timing can help to reduce the congestion and it can improve the traffic flow. We present a new microscopic traffic simulation-optimization model with Particle Swarm Optimization (PSO) for traffic signal timing and demonstrate the model with an application to a major roundabout with 28 traffic signal heads in Izmir, Turkey, using real data. A VISSIM (Verkehr In Stadten-SIMulationsmodell) model of the Osman Kibar Roundabout, based on actual traffic flow data, is developed. A PSO model was used to search for traffic signal timings that would minimize mean travel time through the roundabout, and the microsimulation model (VISSIM) was used to evaluate the solutions from PSO. The changes in the traffic signal timings resulted in a 55.9% decrease in average delay time per vehicle passing through the roundabout, and a 9.3% increase in the number of vehicles passing through the roundabout per unit time in the best-case scenario.