Determinação do impacto de zonas de ultrapassagens proibidas e de faixas adicionais de subida em segmentos de rodovias de pista simples

Abstract

The Highway Capacity Manual (HCM) is generally used to assess capacity and level of service for two-lane rural highways in several countries, including Brazil. However, it was developed with data from North American highways that have different characteristics of vehicles and drivers compared to those found in Brazil. The 6th version of HCM, which has been under final revision, should address capacity and level of service of highways based on Follower Density (FD). This performance measure has been studied and pointed out as more adequate, compared to the performance measures that have been used by the current version of HCM: Average Travel Speed and Percent Time Spent Following, the latter considered unfeasible to be obtained directly from the field. In Brazil, studies have been carried out to obtain a suitable method for calculating level of service on two-lane highways, but there are still few studies that have determined the impact of no-passing zones and climbing lanes, which is the main objective of this research. To achieve this goal, a set of traffic data was obtained from BR-040 highway, in highway segments located in Minas Gerais and Goiás, with the main purpose of calibrating and validating the traffic simulator used in this research, VISSIM. The calibration was performed based on a Genetic Algorithm, but only after estimating other fundamental parameters of the simulator, such as those related to the distribution of desired speeds and the performance of heavy vehicles. With the calibrated simulator, traffic data were generated in hypothetical highway segments with a wide range of geometric and traffic characteristics. Traffic models for such simulation data were adjusted under the following conditions: (i) without no-passing zones and no climbing lanes (called "base models"); (ii) with climbing lanes; and (iii) with nopassing zones. The situations, in terms of geometry and traffic, which determined the greatest impacts due to the presence of climbing lanes and no-passing zones were discussed. Part of the traffic data collected in the field, which were used in the validation of the calibrated simulator, were also used to compare the FD and level of service estimates obtained from the field with other models, including those proposed in this work. The results of this analysis indicated that two quadratic models between directional traffic flow and FD (proposed in this work and another found in the literature) produced level of service and FD values more adherent to those observed in the field. The use of HCM for both the current version and the likely new proposal that will be published in 2019 - without any adaptation - tends to underestimate level of service on Brazilian highways, which would cause investments before the right moment.