Modelos de controle semafórico Fuzzy adaptativo para cruzamentos isolados

Abstract

The traffic of people and goods has direct influence and vital importance in the social and economic development of modern society, and so, its negative consequences originated from the out of order growth deserve attention in the planning of cities. Such mobility demand generates an increase in the flow of vehicles in circulation and consequently, the occurrence of frequent traffic jams, accidents, delays, pollution increase, among others. Several improvements in the traffic control systems have been proposed, and many methodologies have been presented aiming at reducing the effects triggered by those problems. In this work, models are proposed for urban traffic control in isolated crossings through strategiestraced to the timers using fuzzy control and cluster analysis. The objective is to control, adaptively, the flow of vehicles at a crossing by means of a timing plan applied to traffic lights, avoiding or decreasing traffic jams. The proposed models take into account crossings that present Single Exit Condition (CSU), that is, crossings with at least one origin that has only one permissible destination, in contrast to conventional models, which are applicable only to intersections where each origin is associated with more than one destination. The models are composed of three main parts: the first part consists of the estimation of the OD matrix (origin - destination) which is based on Kalman filtering from the traffic count at the crossroads of the study city. This data is used as input to perform the second part, which consists of using clustering techniques to extract the fuzzy sets, and with that, insert them into the fuzzy controller that represents the third part of the methodology proposed here, where the best time for the traffic light is estimated.