Control of Traffic Lights by means of Intelligent Methods

Traffic congestion is a problem in large cities. It has many negative consequences for the economy, the environment and human health. Proper adjustment of the traffic light settings in urban environments is the main means of control of this problem. This paper investigates a network of crossroads situated in central Rome in the area of Municipo V. The study presented in this paper offers alternative approaches to the optimization of signal timings of traffic lights in the network. A comparison is made between the current settings, the Buyanov et al. optimization approach, the nonlinear optimization approach and the bi-level optimization approach. Signal timings for the current settings and for the three optimization approaches are imported in the traffic simulation software suit Aimsun and the output results of the four simulations are compared. The analysis and evaluation of the output data shows that the nonlinear approach gives better results than the Buyanov et al. approach. This is also valid for the bi-level optimization approach. The effectiveness of the used optimization methodologies is graphically presented.