Mathematical Model of reinforcement learning for dynamic traffic light control.

hala Khankhour; Najat  Rafalia; Jaafar  Abouchabaka

doi:10.5269/bspm.79801

hala Khankhour Research laboratory in computer science. Faculty of Sciences, Ibn Tofail University, Kenitra, Morocco
Najat Rafalia
Jaafar Abouchabaka

Résumé

Nowadays, transport has become an essential element for the modern societies. So, the management of networks has also become important. Among the most used tools for the management of these networks, we find traffic lights. These lights do not adapt to the amount of traffic (fixed time for each traffic light). The evolution of new technologies has made it possible to solve this problem and to make traffic lights smart. The objective of this work is to propose a new dynamic control solution for intelligent traffic lights using agent concept and reinforcement learning combined with deep learning. The main advantage of our system is to provide adaptation between traffic lights and smooth traffic flow in different conditions.

Téléchargements

Les données sur le téléchargement ne sont pas encore disponible.

Références

[1] M. Behrisch, L. Bieker, J. Erdmann, & D. Krajzewicz. SUMO–simulation of urban mobility: an overview. In Proceedings of SIMUL 2011, the Third International Conference on Advances in System Simulation. ThinkMind (2011).
[2] J. Wu, A. Abbas-Turki, A. El Moudni, A. "Traffic control in urban intersections using discrete methods." In Proceedings of the 69th Vehicular Technology Conference (VTC Spring) (pp. 1–5). IEEE. (2009).
[3] W. Glankwamdee explored subjects related to Branch and Bound on computational grids in their doctoral thesis at Lehigh University in 2008.
[4] Y. Aibeche. Development of an Intelligent Traffic Light Management System Using Genetic Algorithms (2020).
[5] E. Bingham. Reinforcement learning in neurofuzzy traffic signal control. European Journal of Operational Research, vol. 131, no. 2, p. 232-241, (2001).
[6] W. Gender. Reinforcement Learning for Traffic Signal Control. Doctoral dissertation, McMaster University, Hamilton, Ontario, Canada. Department of Civil Engineering, (2018).
[7] M. Naeem, S. T. H. Rizvi, & A. Coronato. A gentle introduction to reinforcement learning and its application in different fields. IEEE access, vol. 8, p. 209320-209344, (2020).
[8] O. Vinyals, I. Babuschkin, J. Chung, M. Mathieu, M. Jaderberg, W. Czarnecki, & D. Silver. Alphastar: Mastering the real-time strategy game starcraft ii. DeepMind blog, vol. 2, p. 20, (2019).
[9] D. Kalashnikov,A. Irpan, P. Pastor, J. Ibarz, A. Herzog, E. Jang, & S.Levine. Scalable deep reinforcement learning for vision-based robotic manipulation. In Conference on robot learning (pp. 651-673). PMLR, (2018, October).
[10] K.-L. A. Yau, J. Qadir, H. L. Khoo, M. H. Ling, and P. Komisarczuk, "An overview of reinforcement learning models and algorithms applied to the control of traffic signals," ACM Computing Surveys (CSUR), vol. 50, no. 3, p. 34, 2017.
[11] W. Genders and S. Razavi, "Assessment of state representations in reinforcement learning for adaptive traffic signal control," Procedia Computer Science, vol. 130, pp. 26–33, 2018.
[12] J. Gao, Y. Shen, J. Liu, M. Ito, & N. Shiratori. Adaptive traffic signal control: Deep reinforcement learning algorithm with experience replay and target network. arXiv preprint arXiv:1705.02755, (2017).
[13] S. S. Mousavi, M. Schukat, and E. Howley. "Traffic light management employing deep policy-gradient and value-function-based reinforcement learning." In IET Intelligent Transport Systems, volume 11, issue 7, pages 417–423, (2017).
[14] L. Li, Y. Lv, and F.-Y. Wang, in the IEEE/CAA Journal of Automatica Sinica, presented a study on "Traffic Signal Timing Using Deep Reinforcement Learning, vol.3, no. 3 p. 247-254, (2016).
[15] H. Wei, G. Zheng, H. Yao, Z. Li. "Intellilight: A Reinforcement Learning Approach for Intelligent Traffic Light Control." In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, p. 2496-2505, ACM, (2018).
[16] Z. Meng, X. Du, P. Sottovia, D. Foroni, C. Axenie, A. Wieder, C. Sommer., September). Topology-preserving simplification of OpenStreetMap network data for large-scale simulation in SUMO. In SUMO Conference Proceedings (Vol. 3, pp. 181-197), (2022).
[17] D. Dudiak, M. Szabóová, J. Magyar. Exploring Q-Learning in Social Robots for English-Slovak Vocabulary Learning. In 2023 IEEE 23rd International Symposium on Computational Intelligence and Informatics (CINTI), p. 000025-000030, IEEE, (2023, November).
[18] Z. Zhang. Introducing Reinforcement Learning to High School Students Through the Integration of Physical Robots and Virtual Interfaces (Doctoral dissertation, Tufts University), (2023).
[19] H. Chaudhuri, V. Masti, V. Veerendranath, S. Natarajan. A comparative study of algorithms for intelligent traffic signal control. In Machine Learning and Autonomous Systems: Proceedings of ICMLAS 202, p. 271-287, Singapore: Springer Nature Singapore, (2022).