The impending widespread availability of realtime traffic condition data from vehicles serving as traffic probes offers the possibility of significant advances in traffic signal control performance. In order to realize these advances, it is necessary to understand the capabilities and limitations of traffic probe data collection. In this paper, vehicle trajectories from a microscopic simulation of an arterial corridor are used as the baseline truth model, and are then sampled according to a variety of sampling strategies that could be implemented on probe vehicles. Measures of effectiveness relevant to traffic signal control, such as travel times and speeds, queue lengths and percentages of vehicles that need to stop, are estimated using the sampled data and these estimates are compared with the baseline truth values. These comparisons show the difficulty of estimating the MOEs at low market penetrations of probe vehicles, indicating that at the lower market penetrations it will probably be necessary to aggregate data from multiple signal cycles in order to obtain reliable enough outputs to use for real-time signal control adjustments.
Evaluation of probe vehicle sampling strategies for traffic signal control
2011-10-01
912252 byte
Conference paper
Electronic Resource
English
A Probe-Vehicle-Based Evaluation of Adaptive Traffic Signal Control
Online Contents | 2012
|Probe Vehicle Sampling for Real-Time Traffic Data Collection
British Library Conference Proceedings | 2005
|Adaptive traffic signal control algorithms based on probe vehicle data
Taylor & Francis Verlag | 2021
|