Application of a neuro fuzzy network to forecast the behavior of the urban traffic of the city of São Paulo

In recent years the behavior of the parcel delivery market has undergone significant changes influenced by increased competitiveness, concern with the environment, electronic commerce, the increase in the number of customers, reducing the size of applications; expansion of the collection sites and distribution, increasing the flow of urban vehicles, increasing fleet; complexity of the operations of collection and distribution, and increased restrictions of movement of vehicles, increased restrictions on size and capacity of vehicles. Thus the routing and scheduling of vehicles tend to receive a focused mission, observing the requirements of the consumer market and end customer within the urban distribution logistics. The aim of this paper is to apply a neuro fuzzy network to forecast the behavior of the urban traffic of the city of São Paulo aided routing in a hierarchy of three levels: strategic represented by forecasting the behavior of traffic, represented by the tactical and operational routing dynamics represented by routing algorithms. The methodology of the work consists in the capture of relevant events that affect the flow of traffic of the city of São Paulo and implementation of a neuro fuzzy network trained with these events to predict the behavior of traffic. This paper will present the partial results of the use of neuro fuzzy network in predicting the behavior of the traffic of the city of São Paulo.