Cycle-to-cycle queue length estimation from connected vehicles with filtering on primary parameters

Comert, Gurcan; Begashaw, Negash · 2021 · OpenAlex-citations

DOI: 10.1016/j.ijtst.2021.04.009

archive: archived pipeline: cataloged verified

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

Summary

This study addresses the challenge of accurate cycle-to-cycle queue length estimation (QLE) at traffic intersections using connected vehicle (CV) data, particularly under low market penetration rates (MPR). Existing QLE models often assume known arrival rates and MPRs or estimate them in real-time; however, these estimators produce significant errors at low MPRs, rendering them inefficient for traffic control applications. The authors propose using filtering algorithms—specifically Kalman filters (KF) and Particle filters (PF)—to estimate primary parameters (arrival rate and MPR) more accurately, thereby improving the overall QLE accuracy without requiring known ground-truth parameters. The methodology employs microsimulations using the Vissim software to generate vehicle trace files for an isolated intersection with a 90-second signal cycle. The study evaluates estimator performance across various volume-to-capacity ratios (0.60 to 0.99) and MPRs (0.1% to 90%). To simulate dynamic conditions, arrival rates and MPRs were changed every 15 minutes. The filtering algorithms were applied to estimate the primary parameters from raw CV data (location, time, and count), which were then used in QLE equations. The study compares the performance of filtered estimators against raw CV estimators and scenarios where primary parameters were known. The results demonstrate that both Kalman and Particle filters effectively converge to true parameter values within 15 minutes, even when parameters change dynamically. At low MPRs (below 20%), filtering significantly reduces estimation errors compared to unfiltered methods. Specifically, at a 20% MPR, filtered estimators met or surpassed the accuracy of scenarios with known parameters. The study found that filtering is particularly beneficial for high volume-to-capacity ratios, where queue fluctuations are significant. Additionally, using the last known estimated queue length when no connected vehicles are present yielded better results than inputting average estimated values. Both filtering algorithms proved computationally efficient, requiring less than 0.1 seconds per cycle, making them suitable for real-time traffic operations. The significance of this work lies in its ability to provide accurate QLEs under low market penetration conditions, which are common in rural areas or during early CV adoption phases. By improving the estimation of primary parameters through filtering, the method reduces the reliance on high MPRs or extensive sensor infrastructure. This approach facilitates the deployment of intelligent transportation system controls in environments with limited CV coverage, ensuring robust performance even with sparse data. The findings suggest that filtering primary parameters is a viable strategy for enhancing the reliability of CV-based traffic management systems.

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.

• traffic density