Microscopic Estimation of Freeway Vehicle Positions From the Behavior of Connected Vehicles

Goodall, Noah J.; Smith, Brian L.; Park, Byungkyu · 2014 · OpenAlex-citations

DOI: 10.1080/15472450.2014.889926

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of limited data availability in connected vehicle (CV) systems, where only a fraction of vehicles transmit status data due to bandwidth constraints and slow deployment. While CV applications like adaptive signal control and ramp metering require high penetration rates (often 20–30% or more) to function effectively, the authors propose a method to artificially augment these rates by estimating the positions of non-communicating ("unequipped") vehicles. The motivation is to enable traffic management benefits during the transitional period when only a small percentage of vehicles are equipped with communication technology. The authors developed an algorithm that predicts the locations of unequipped vehicles by analyzing the behavior of nearby equipped vehicles. The method relies on the Wiedemann car-following model, which predicts expected vehicle acceleration based on relative speed and headway. When an equipped vehicle’s actual acceleration deviates significantly from the model’s prediction (specifically, when it decelerates more than expected), the algorithm infers the presence of an unseen leading vehicle. It then calculates the unequipped vehicle’s position and speed using empirical relationships and inserts this "estimated vehicle" into the traffic simulation. The algorithm assumes equipped vehicles report location, speed, and acceleration once per second. The algorithm was evaluated using high-resolution trajectory data from the Next-Generation Simulation (NGSIM) project, specifically a 500-meter section of Interstate 80. The study introduced a metric called "effective penetration rate" to measure accuracy, pairing estimated positions with observed unequipped vehicles. Results indicated that with only 10% of vehicles communicating, the algorithm could predict the locations of 30% of total vehicles with 9-meter accuracy. Performance was strongest during or downstream of congestion, as the method relies on vehicle interactions. At high penetration rates (above 80%), the algorithm tended to overestimate densities due to fewer unequipped vehicles to detect. To demonstrate practical utility, the algorithm was integrated with the GAP ramp metering algorithm, which controls on-ramp signals based on predicted gaps in mainline traffic. Testing showed that adding the location estimation algorithm significantly improved ramp metering performance at low CV penetration rates (10% and 25%), reducing vehicle delay and increasing network speed compared to using CV data alone. At higher penetration rates, the algorithm maintained performance without degradation. The findings suggest that microscopic estimation of unequipped vehicle positions can extend the benefits of connected vehicle applications to environments with low equipment penetration, particularly in congested conditions.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• v2x connected vehicle