Characteristics of Transitions in Freeway Traffic

This proposed work seeks to understand the characteristics of transitions as freeway traffic changes from one state to another. Transitions occur gradually over time and space, and their temporal and spatial features are relatively unknown. The work will address the features of three different types of transitions which occur near 1) the tail of a queue 2) the head of a queue (i.e. an active bottleneck) and 3) inhomogeneous points on freeway (e.g. merges and diverges). The transition zone near the tail of a queue moves upstream as the queue grows, and moves downstream as the queue recedes with decreasing…

This proposed work seeks to understand the characteristics of transitions as freeway traffic changes from one state to another.  Transitions occur gradually over time and space, and their temporal and spatial features are relatively unknown.  The work will address the features of three different types of transitions which occur near 1) the tail of a queue 2) the head of a queue (i.e. an active bottleneck) and 3) inhomogeneous points on freeway (e.g. merges and diverges).  The transition zone near the tail of a queue moves upstream as the queue grows, and moves downstream as the queue recedes with decreasing demand.  Hence, the dynamics of the transition zone will be explored by analyzing the relationship between the duration of transition (at a fixed location) and various traffic and location variables (e.g. distance from the bottleneck, change in flow before and after a regime change, etc).  The second and third types of transitions are stationary such that they do not propagate over space as a shock wave.  For these types of transitions, the length of the transition zones with respect to traffic and location variables will be analyzed. Several U.S. and non-U.S. freeway sites will be selected for this study in order to verify reproducibility and compare differences across sites.  In particular, researchers will use data from inductive loop detectors for the analyses of transition zones near the tails of queues.  These detector data are suitable for analyzing this type of transition since the propagation of a transition zone can be observed over a long distance.  The length of transition can be estimated based on the duration observed at a detector location.  For the other two types of transitions, datasets from the Next Generation Simulation will be utilized.  These datasets provide individual vehicle trajectories whose resolution is suitable for analyzing these types of transitions.  The length of a transition zone will be measured directly from the vehicle trajectories. This research will provide a valuable insight on how congested traffic behaves under various transitions that frequently occur on urban freeways.  Hence, the results will expand the current knowledge on traffic congestion and serve as a building block for future traffic modeling and management practice.