Agent-based simulation to assess the performance of intersections with pre-signals: comparison with roundabouts

Abstract

Solving traffic congestion problems in intersections is a complex logistic problem that usually consists in building infrastructures, such as bridges, tunnels, or roundabouts, which represent the costliest solutions. Concerning the case of roundabouts, many parameters influence its performance, e.g. geometry and size. Thus, other less costlier solutions should be pondered. This paper introduces a low-cost approach to traffic intersections, by using pre-signals; and conducts simulation experiments to verify if this approach could be used to improve the performance of traffic intersections. In this sense, an agent-based traffic simulation model was developed that applies the object modelling paradigm of Simio to model the individual behaviour of vehicles. The simulation experiments results indicate that the flow of vehicles can be increased up to 20%, reducing the average queue sizes and crossing time per vehicle and saving the fuel consumed up to 64%.