An agent-based simulation model of dynamic real-time traffic signal controller

Aljaafreh, Ahmad; Saleh, Maen; Al-Oudat, Naeem; Alaqtash, Murad · 2020 · Crossref

DOI: 10.19101/ijacr.2019.940085

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 minimizing driver waiting times and avoiding traffic jams at signalized intersections by proposing an adaptive, real-time traffic signal controller. While conventional fixed-time controllers are simple, they fail to adapt to variable traffic conditions, leading to inefficiencies. The authors aim to overcome this by developing a multi-agent system that dynamically adjusts green light durations based on real-time traffic density and queue lengths, thereby optimizing the average waiting time (AWT) for vehicles. The researchers modeled and simulated the controller using Matlab/Simulink/SimEvents for a single four-way intersection. The system employs an agent-based methodology decomposed into five sub-agents: vehicle, queue, server, controller, and roadside unit (RSU). A key innovation is the elimination of direct inter-vehicle communication; instead, vehicles report their position and speed to RSUs, which monitor virtual road segments. The controller calculates traffic density and queue size to determine the green time for each approach. The simulation assumed vehicle speeds uniformly distributed between 20 and 60 km/h, with inter-arrival times following an exponential distribution. The model was verified through 50 runs of 10,000 simulation time units, analyzing parameters such as green phase duration, service time, and AWT. The results demonstrate that the proposed adaptive controller outperforms a fixed-time controller when there is significant variance (variance > 10) in the mean inter-arrival times of approaching vehicles. Specifically, as the variance in inter-arrival times increases, the adaptive controller maintains lower average waiting times compared to the fixed-time baseline. Conversely, the fixed-time controller performs better when variance is low (less than 10). The study also identified an optimal cycle time of approximately 50 seconds, regardless of inter-arrival time variance. Verification simulations confirmed that the model correctly handles random service times and adjusts green phases equally when traffic density is invariant. The significance of this work lies in providing a robust, agent-based framework for intelligent transportation systems that reduces fuel consumption and emissions by minimizing delays. By utilizing RSUs and virtual road segmentation, the model offers a practical solution that avoids the complexities of direct vehicle-to-vehicle communication. The findings suggest that adaptive controllers are superior in dynamic traffic environments, though they require more complex implementation than fixed-time systems. Future work aims to extend the model to distributed control systems involving multiple intersections and to adapt cycle lengths and phase sequences dynamically.

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.

• mental model of traffic

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Theoretical Contribution: computational model