Travel Time Prediction-Based Routing Algorithms for Automated Highway Systems

Abstract

This study investigates routing algorithms for automated highway systems (AHS). In AHS, the central system manages decisions regarding routing for all vehicles and the distribution of traffic volume. We define an automated highway routing problem, of which the objective is to minimize the average travel time of vehicles through the target highway network. We propose four routing approaches considering (1) distance, (2) current traffic conditions, (3) predicted travel time, and (4) probabilistic route selection with predicted travel time. In the third and fourth approaches, the predicted travel time is obtained from an empirical speed-density relationship. AnyLogic, an agent-based simulation software, is used to simulate the behavior of individual cars. Four approaches are tested on a sample highway network and we found that the routing approach considering the predicted travel time difference exhibits the best performance.