Modelagem matemática para otimização do tráfego urbano semaforizado

Abstract

Due to the growing increase of the vehicle fleet in most cities, the traffic systems present in their streets end up not to permit the large number of existing vehicles, emerging inevitable traffic jams and, consequently, several problems related to them. Considering the importance of the use of mathematical optimization techniques to solve real problems, this paper deals with the investigation and proposal of use of these techniques in programming traffic lights in order to find optimal times, which considerably reduce the time of vehicular travel in cities of medium and large size. This work presents two ptimization models which, implemented and considering the region traffic data intended to optimize, enable to provide an automated way of green light timing that aim to minimize the time ofvehicular travel. Also, it presents the computational results obtained in the investigation of these models in two theoretical examples, one based on data available in the literature and the other based on collected data related to the traffic of a central region of the city of Passo Fundo, RS. In the first example, a comparison was made between the results obtained by the models and those provided by the Transyt 7F software; whereas in the second example, the performance of vehicular traffic belonging to the streets before and after the process of optimization are compared through theoretical index. The results were positive in both instances, giving support to a possible use of the models in a real situation