Development of an Appropriate Model for the Optional Lane-changing Rate in Weaving Segments with Short Lengths

Abstract

In order to compute level of service and density in weaving segments, the Highway Capacity Manual (HCM) defined for the first time in 2010 a relationship based on a lane change rate to assess the density of the weaving segment. It is critical to accurately estimate lane changing rate in these situations, but field observations in weaving segments shorter than 250 meters in Tehran, Iran revealed a significant difference between the HCM2016 model estimate and field data. The traffic and geometric data collected at 87 (15-minute) intervals from six weaving segments in Tehran were used to develop models for estimating lane changing rate in weaving segments. These 87 intervals were then divided into 69 (terrain data) for equations and 18 (test data) for model comparisons. Weaving volume and weaving segment area are introduced as two independent variables in the optional lane changing rate model of weaving vehicles in this study, with R2=0.74. Furthermore, for a lane-changing model of non-weaving vehicles with R2=0.95, two new variables of non-weaving volume and traffic solidity were defined. Finally, based on the 18 intervals used to test the results, it showed the improvement of the developed models results compared to HCM2016 models.