Controlling Traffic Jams on Urban Roads Modeled in Coloured Petri Net using Genetic Algorithm

Abstract

This paper presents an application to traffic lights control in congested urban traffic, in real time, taking as input the position and route of the vehicles in the involved areas. This data is obtained from the communication between vehicles and infrastructure (V2I). Due to the great complexity of the possible combination of traffic lights and the short time to get a response, Genetic Algorithm was used to optimize this control. According to test results, the application can reduce the number of vehicles in congested areas, even with the entry of vehicles that previously were not being considered in these roads, such as parked vehicles.