Dynamic data collection of staggered-following behavior in non-lane-based traffic streams

Abstract

Different maneuvering patterns of vehicles, absence of lane discipline and interactions of a large number of vehicles with each other and with roadway features make the traffic phenomena of non-lane-based traffic streams more complex. Vehicles’ movement in weak lane discipline traffic is rather two-dimensional because they always tend to evaluate possible available gaps on the road while progressing longitudinally. Recent literature has underlined the importance of centerline separation of traffic in modeling the staggered-following behavior. However, the requirements for staggered-following trajectory data are indeed stringent. Although recent advancements in new digital technology have expedited new horizons in the field of traffic engineering, proper estimation of microscopic car-following data has still proven to be challenging. Understanding that a proper evaluation of car-following behavior in non-lane-based traffic environments requires an accurate characterization of the microscopic traffic variables and reliable experimental data, this study describes an image-based in-vehicle trajectory data collection system to process the microscopic variables (such as longitudinal gap, centerline separation, vehicle speeds and accelerations), using camera calibration and in-vehicle GPS information on straight roads. Improved accuracy in the experimental data collection, proper extraction and estimation of data can substantially enrich the understanding of riders’ behavioral phenomena from a microscopic perspective and the realism of traffic sub-models, which will result in a better prediction of microscopic simulation models.