Reinforcement Learning for Dynamic Traffic Signal Control Using Advanced Machine Learning Techniques

References

[1] F. Rasheed, Q. Ni, and H. Xu, “DRLfor Vehicle Actuated Control (VAC): A review,” IEEE Access, vol. 8, pp. 208016–208044, 2020.

[2] D. Ma, Y. Wang, Q. Guo, and Y. Bie, “A DRL approach to Vehicle Actuated Control (VAC) with temporal traffic pattern mining,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 8, pp. 11789–11800, Aug. 2021.

[3] K. J. Prabuchandran, H. Kumar A. N., and S. Bhatnagar, “Multi-agent RL for Vehicle Actuated Control (VAC),” in Proc. 17th IEEE Int. Conf. Intelligent Transportation Systems (ITSC), Qingdao, China, 2014, pp. 2292–2297.

[4] N. S. Jadhao and A. S. Jadhao, “Vehicle Actuated Control (VAC) using RL,” in Proc. 4th Int. Conf. Communication Systems and Network Technologies, Bhopal, India, 2014, pp. 273–277.

[5] L. Li, Y. Lv, and F.-Y. Wang, “Traffic signal timing via deep RL,” IEEE/CAA Journal of Automatica Sinica, vol. 3, no. 3, pp. 247–254, Jul. 2016.

[6] S. M. A. Shabestary and B. Abdulhai, “Deep learning vs. discrete RL for adaptive Vehicle Actuated Control (VAC),” in Proc. 21st IEEE Int. Conf. Intelligent Transportation Systems (ITSC), Maui, HI, USA, 2018, pp. 286–293.

[7] K.-F. Chu, A. Y. S. Lam, and V. O. K. Li, “Vehicle Actuated Control (VAC) using end-to-end off-policy deep RL,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 7, pp. 7184–7195, Jul. 2021.

[8] T. Chu, J. Wang, L. Codecà, and Z. Li, “Multi-agent DRLfor large-scale Vehicle Actuated Control (VAC),” IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 3, pp. 1086–1095, Mar. 2019.

[9] L. A. Prashanth and S. Bhatnagar, “RL with function approximation for Vehicle Actuated Control (VAC),” IEEE Transactions on Intelligent Transportation Systems, vol. 12, no. 2, pp. 412–421, Jun. 2010.

[10] W. Liu, Y. Zheng, S. Zhou, and J. Wang, “Distributed cooperative RL-based Vehicle Actuated Control (VAC) that integrates V2X networks’ dynamic clustering,” IEEE Transactions on Vehicular Technology, vol. 66, no. 10, pp. 8667–8681, Oct. 2017.

[11] Y. Dai, D. Zhao, and J. Yi, “A comparative study of urban Vehicle Actuated Control (VAC) with RL and adaptive dynamic programming,” in Proc. Int. Joint Conf. Neural Networks (IJCNN), Barcelona, Spain, 2010, pp. 1–8.

[12] R. Zhang, A. Ishikawa, W. Wang, B. Striner, and O. K. Tonguz, “Using RL with partial vehicle detection for intelligent Vehicle Actuated Control (VAC),” IEEE Transactions on Intelligent Transportation Systems, vol. 22, no. 1, pp. 404–415, Jan. 2020.

[13] X. Wang, L. Ke, Z. Qiao, and J. Ye, “Large-scale Vehicle Actuated Control (VAC) using a novel multiagent RL,” IEEE Transactions on Cybernetics, vol. 51, no. 1, pp. 174–187, Jan. 2021.

[14] N. Kodama, T. Harada, and K. Miyazaki, “Vehicle Actuated Control (VAC) system using DRLwith emphasis on reinforcing successful experiences,” IEEE Access, vol. 10, pp. 128943–128950, 2022.

[15] T. Wu, P. Zhou, K. Liu, and Y. Wang, “Multi-agent DRLfor urban traffic light control in vehicular networks,” IEEE Transactions on Vehicular Technology, vol. 69, no. 8, pp. 8243–8256, Aug. 2020.

[16] Y. Li, Z. Wang, X. Chen, and F.-Y. Wang, “Regional multi-agent cooperative RL for city-level traffic grid signal control,” IEEE/CAA Journal of Automatica Sinica, vol. 11, no. 9, pp. 1987–1998, 2024.

@article{1717642,
      author = {Ganapathi Sai Abhishek Grandhe, Dr. Gnana Prakasam Thangvel},
      title = {Reinforcement Learning for Dynamic Traffic Signal Control Using Advanced Machine Learning Techniques},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {9},
      number = {11},
      pages = {1490-1496},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1717642.pdf},
      abstract = {Traffic jam in cities has become one of the major issues of modern cities due to high rates of urbanization, rise in the number of vehicles and poor development of road network. Conventional traffic light systems are normally based on fixed time or rule-based control measures that are not able to be responsive to changing and unanticipated traffic conditions. These systems therefore have a tendency of increasing the waiting time of vehicles, the length of queues, and inefficient use of the road capacity. RL is now a promising technology that can be utilized to develop the intelligent and adaptive Vehicle Actuated Control (VAC) systems that will be able to learn the best control policies by continually interacting with the traffic environment. In this work a RL based dynamic TSC framework has been Suggested with the Vehicle Actuated Control (VAC)ler being an intelligent agent, which monitors the real time traffic such as the length of queues, the arrivals made by various vehicles and the current signal phase. The agent, with assistance of these observations, selects the appropriate signal actions to maximize traffic flow and to reduce congestion. PPO and DQN are DRL algorithms, which are used to learn effective signal control strategies through training in a simulation. The effectiveness of the Suggested system is probed and contrasted to the conventional fixed-time traffic lights control systems. The RL-based controller is indeed capable of causing a dramatic decrease in the average length of queues, decreased waiting time of vehicles and rised traffic throughput, as the results of the experiment under consideration demonstrate. The suggested solution may be seen as a scale-up and adaptive solution to smart transportation networks and contributes to the creation of effective traffic control in smart cities.},
      keywords = {Intelligent Transportation Systems, Multi Agents Systems, Traffic Optimization, Machine Learning, Smart Cities.},
      month = {May},
      doi = {https://doi.org/10.64388/IREV9I11-1717642}
  }