Estimação da brecha crítica em interseções não semaforizadas utilizando simulação microscópica de tráfego

• Ano de defesa: 2018
• Autor(a) principal: Araújo, Alessandro Macêdo de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Transportes; Simulação de tráfego; Calibração; Microscopic simulation; Calibration; TransModeler
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/37433

One of the main points in the modeling of urban traffic is to represent the formation and discharge of queues at the intersections, since they are responsible for most of the impedances suffered by the users in the road network. Microsimulation is one of the approaches to modeling traffic and it is a tool that is increasingly being used. The calibration of microsimulation models is usually performed based on the traffic performance measures to be estimated, which depend on several factors that, when not previously controlled, can lead to inadequate adjustments to the calibration parameters. The main objective of this study is to propose a method to estimate critical gap in urban two-way stop-controlled intersections using microscopic simulation. The proposed method is based on the time that the driver takes in the first position of the queue while waiting for an acceptable gap, being also suggested the verification of the major road’s arrival pattern through hypothesis test on headways distribution. To estimate the critical gap, both 2 and the difference between the waiting times’ means (simulated and observed) were used as goodness of fit criteria; the 2 test and the confidence interval of the difference between the means were used to validate the calibrated parameter value. The method also included an adjustment in the simulated waiting times to better match observed field behavior. The case study is composed of two intersections located in the city of Fortaleza, with very different levels of demand. The results indicated minimum gap time values higher than the default value of 3.0 seconds, thus suggesting that the average profile of the drivers at the study sites is a bit more cautious than the standard VISSIM profile; in the case of results obtained in TransModeler, the critical headway was generally lower than the default mean value. For significantly different traffic demands, the same calibrated value was not adequate in most cases. Another objective of this study is to compare the proposed calibration procedure with the conventional one, in which the calibration target measure is a traffic measure of effectiveness, such as the average queue length. The proposed method yielded more accurate results.