Modeling of Effective Parameters for Capacity Prediction at Signalized Intersection Lanes

Aydın, Metin Mutlu · 2022 · OpenAlex-citations

DOI: 10.7250/bjrbe.2022-17.578

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 inadequacy of current capacity manuals, such as the Highway Capacity Manual (HCM), in accurately predicting lane capacity at signalized intersections in developing countries. Existing models often fail to account for local irregularities caused by undisciplined vehicle movements and lane utilization, such as roadside parking blockages, failure to obey distance rules, and the illegal formation of extra lanes. These factors significantly reduce actual capacity compared to theoretical predictions. To resolve this, the research aims to develop lane-based capacity estimation models that incorporate these effective parameters, specifically tailored to local traffic conditions in Turkey. The methodology involved a comprehensive data collection process at six signalized intersections (including four-legged, four-legged roundabout, and three-legged roundabout types) in Antalya and Trabzon, Turkey. Data were gathered during peak hours in both winter and summer seasons using video recordings and site observations to ensure natural driver behavior. The dataset included 56,760 vehicles, with detailed metrics such as traffic volume, number of actively used lanes, vehicle queue lengths, signal timing parameters, and vehicle composition. Lane capacity and saturation flow rates were calculated using HCM equations. To model the impact of various parameters on capacity, the study employed two distinct methods: Ordinary Least Squares (OLS) regression and a swarm intelligence-based metaheuristic search algorithm known as the Artificial Bee Colony (ABC) algorithm. The results demonstrated a strong correlation between measured lane volume and calculated capacity, supporting the feasibility of a lane-based model. The study developed two new estimation models, ALLCEM-1 and ALLCEM-2, using the statistical and metaheuristic approaches, respectively. Both methods proved effective in modeling lane capacity. The analysis identified several key parameters that significantly impact prediction accuracy: intersection type (roundabout vs. non-roundabout), effective green time, saturation flow rate, traffic volume, heavy vehicle ratio, and the number of actively used lanes. The findings highlight that irregular lane utilization, such as illegal parking or extra lane formation, creates measurable discrepancies between designed and actual capacity. The significance of this research lies in its provision of localized capacity estimation tools for developing nations where standard manuals may not reflect local driving behaviors and traffic chaos. By integrating parameters related to undisciplined movements and lane utilization, the proposed ALLCEM models offer a more accurate assessment of intersection performance. This allows transportation engineers and planners to better design signal timings and geometric properties to mitigate congestion, delays, and accidents. The study underscores the necessity for countries to develop or adapt capacity models that reflect their specific traffic characteristics rather than relying solely on manuals developed for different contexts.

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