Authors : Rupali Shankar Kale; Dr. Manisha Bharati

Volume/Issue : Volume 11 - 2026, Issue 5 - May

Google Scholar : https://tinyurl.com/4nd8ab76

Scribd : https://tinyurl.com/smfdnbtj

DOI : https://doi.org/10.38124/ijisrt/26May1200

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : Urban traffic congestion is a major challenge due to increasing vehicle density and the limitations of conventional fixed time traffic signal systems. Traditional signal controllers lack adaptability and fail to respond effectively to dynamic traffic conditions, especially in heterogeneous urban traffic environments such as Indian roads. This research proposes an intelligent adaptive traffic signal control framework integrating Deep Reinforcement Learning (DRL) with computer visionbased traffic perception. The proposed system employs Proximal Policy Optimization (PPO) as the primary learning algorithm and compares its performance with Deep Q-Network (DQN) and Advantage Actor-Critic (A2C). Traffic flow is simulated using the SUMO traffic simulator and controlled in real time through the TraCI interface. Vehicle density, lane occupancy, and congestion states are extracted using YOLOv8-based perception from rendered simulation frames. Experimental analysis demonstrates that the PPO-based adaptive controller significantly reduces average waiting time and queue length while improving traffic throughput compared to conventional fixed time systems and other reinforcement learning approaches.

Keywords : Traffic Signal Control, Reinforcement Learning, PPO, YOLOv8, SUMO, TraCI, Intelligent Transportation Systems, Deep Learning

References :

1. L. Li, Y. Lv, and F.-Y. Wang, “Traffic signal timing via deep reinforcement learning,” IEEE/CAA Journal of Automatica Sinica, vol. 3, no. 3, pp. 247–254, 2016.
2. H. Wei et al., “IntelliLight: A reinforcement learning approach for intelligent traffic signal control,” in Proc. 24th ACM SIGKDD Int. Conf. Knowl. Discovery Data Mining (KDD), 2018, pp. 2496–2505.
3. I. Arel, C. Liu, T. Urbanik, and A. G. Kohls, “Reinforcement learning-based multi-agent system for network traffic signal control,” IET Intelligent Transport Systems, vol. 4, no. 2, pp. 128–135, 2010.
4. E. Van der Pol and F. A. Oliehoek, “Coordinated deep reinforcement learners for traffic light control,” in Proc. 15th Int. Conf. Autonomous Agents and Multiagent Systems (AAMAS), 2016.
5. J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, “Proximal policy optimization algorithms,” arXiv preprint arXiv: 1707.06347, 2017.
6. G. Zheng, Y. Wang, and X. Wang, “Learning phase competition for traffic signal control,” in Proc. 29th ACM Int. Conf. Information and Knowledge Management (CIKM), 2020, pp. 1965–1974.
7. T. Chen et al., “PPO-based multi-intersection traffic signal control with graph attention networks,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 11, pp. 12345–12356, 2022.
8. Q. Guo et al., “Vision-based reinforcement learning for adaptive traffic signal control using YOLO and deep RL,” Expert Systems with Applications, vol. 215, 2023.
9. X. Liang et al., “Deep reinforcement learning for traffic signal control with computer vision,” Transportation Research Part C: Emerging Technologies, vol. 128, 2021.
10. W. Genders and S. Razavi, “Using a deep reinforcement learning agent for traffic signal control,” arXiv preprint arXiv: 1611.01142, 2016.
11. V. Mnih et al., “Asynchronous methods for deep reinforcement learning,” in Proc. 33rd Int. Conf. Machine Learning (ICML), 2016, pp. 1928–1937.
12. Ultralytics, “YOLOv8 Documentation,” 2023. [Online].  Available: https://docs.ultralytics.com.
13. P. A. Lopez et al., “Microscopic traffic simulation using SUMO,” in Proc. 21st IEEE Int. Conf. Intelligent Transportation Systems (ITSC), 2018, pp. 2575–2582.
14. Z. Wu et al., “YOLOv8-based real-time vehicle detection for intelligent traffic management,” Sensors, vol. 24, no. 5, 2024.