Análise de modelos simplificados para determinação do fluxo de saturação em interseções semaforizadas da cidade de Toledo-PR

Abstract

Regarding traffic lights planning projects, no matter the method, the obtainment the saturation flow ratio is required (FS), in order to define the green time. The DENATRAN manual indicates the obtaining of this parameter through vehicle-counting and frequency histogram analysis. However, this method may be costly. Thus, it is verified the existence of a range of methods, both national and international ones, which aims to estimate FS. Generally, these methods have been developed in large cities with high vehicle flows. In this way, there was an analysis of the main simplified models in order to estimate the FS in intersections with traffic lights in Toledo, Paraná. In total, it were evaluated five national and eight international models. For the validation of these models, the results were compared to values obtained from the Shanteau’s cumulative curve, which consists of the analysis of onsite video records. The models that presented a relative-residue lower than 20% were validated. In total, seven out of thirteen models analyzed were considered applicable for the FS data collection at the city of Toledo-PR. In addition, three different models were generated from multiple regression resulting from field data, which resulted in a relative residue lower than 13%. Nonetheless, all the models have presented a determination coefficient lower than 50%. In a general way, it is verified that the saturation flow values are influenced by the driver’s behavior, especially because of line vehicles close proximity (gap), and speed and, therefore, resulting in the obtainment of a low determination coefficient. The research also allowed the analysis of the passenger car equivalent factor (double saturated headway analysis) and the initial time loss (Shanteau’s cumulative curve) onsite.