抄録
The present study has been carried out for examining travel time reliability under heterogeneous traffic conditions on three different urban arterial road sections in the cities of Ahmedabad and Surat in India. Travel time reliability indices along with descriptive statistics over selected periods are calculated and a new reliability measure called Reliable Buffer Index (RBI) is proposed in this research work for evaluating the performance of selected urban arterial road sections. This new measure has the credibility of explaining inherent causes and factors that affect travel time variations in short time intervals for a given traffic volume. The study also focuses on the determination of best fit or potential statistical distributions for analyzing travel time variations using car travel time data. The study also highlights the development of the reliability-based Level of Service (LOS) thresholds on the basis of variation in identified effective reliability measures with respect to coefficient of variation using k-means clustering technique. The study results indicated that Buffer Time (BT) and Buffer Time Index (BTI) are the most effective measures that can capture the travel time variations. The study also presents a brief discussion on applications and limitations of existing travel time reliability measures that are commonly used. From the study results, it was observed that Burr or lognormal distributions are the potential distributions for modeling car travel times under heterogeneous traffic conditions. The methodology and study results can be helpful for transportation planners in comparing the performance or efficiency of different road facilities in a network in terms of travel time reliability in developing countries like India.
