PDF

Description

The rise of mobile internet has changed routing behavior in traffic networks. With ubiquity of traffic information and the increased use of routing apps, urban and suburban areas in the US and abroad have seen a recent rise in "cut-through" traffic and related congestion patterns. The phenomenon is shown to have negative impact on communities near congested major roads. Chapter 1 presents data that shows the rerouting phenomenon from highway to side-roads on the I-210 corridor in the LA Basin.

The differences in the routing behavior of routing-app users and non routing-app users remains uncertain to this day. Therefore, the impact of these applications on traffic is unclear. Chapter 2 introduces three traffic assignment models that differentiate app users' and non-app users' routing behaviors. Two of them are based on static traffic assignment models - the so called cognitive cost path choice model and the restricted path choice model - and the third one on dynamic traffic assignment models using the microsimulator Aimsun.

Chapter 3 provides a criterion to evaluate the impact of routing apps usage on road traffic at a macroscopic level: the average marginal regret. Derived from game theory, the average marginal regret of an observed state of traffic can be seen as a distance between this observed state and a user equilibrium state of traffic (the so called Nash equilibrium for routing in traffic networks). Experiments - using the two previously introduced static models - demonstrate that the average marginal regret decreases with an increase of app usage. Similar results are shown using the dynamic model in Aimsun. A sensitivity analysis of the restricted path choice model equilibrium with respect to the app usage ratio even proves that the average marginal regret monotonically converges to zero with an increase of app usage. Therefore, chapter 3 shows that an increase of app usage stirs a state of traffic toward a user equilibrium state. This is plausible as one can expect such property. Recall that a user equilibrium is most likely not socially optimal. Therefore, app usage might leads to an increase of the average travel time in the network at a system level.

Details

Files

Statistics

from
to
Export
Download Full History