Abstract
Reinforcement Learning knows an important success in various applications such as robotics, games, resource management, etc. However, it is proving insufficient to solve the problem of crowd evacuation, in a realistic environment because the crowd situation is very dynamic, with many changing variables and complex constraints that make it difficult to solve. And there is no standard reference environment that can be used to train agents in an evacuation. A realistic environment can be complex to design. In this paper, we use Deep Reinforcement Learning to train agents in evacuation planning. The environment is modeled as a grid with obstacles and the solution is modeled using intelligent agents. It takes into account certain parameters, such as the number of occupants, the capacity level, and the time to pass through the exit doors. The objective is to allow the evacuation of the occupants as quickly as possible and to help the agent decide on the optimal escape route under varying conditions over time. And subsequently, this approach will be useful to evacuation decision makers to better implement dynamic arrow signs. The results are motivating to use this type of learning to optimize decision support in evacuation situations.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ciaburro, G.: Keras reinforcement learning projects: 9 projects exploring popular reinforcement learning techniques to build self-learning agents. Packt Publishing Ltd. (2018)
François-Lavet, V., Henderson, P., Islam, R., Bellemare, M.G., Pineau, J.: An introduction to deep reinforcement learning. arXiv preprint arXiv:1811.12560 (2018)
Kaelbling, L.P., Littman, M.L., Moore, A.W.: Reinforcement learning: a survey. J. Artif. Intell. Res. 4, 237–285 (1996)
Li, X., Liang, Y., Zhao, M., Wang, C., Bai, H., Jiang, Y.: Simulation of evacuating crowd based on deep learning and social force model. IEEE Access 7, 155361–155371 (2019)
Martinez-Gil, F., Lozano, M., Fernández, F.: Multi-agent reinforcement learning for simulating pedestrian navigation. In: Vrancx, P., Knudson, M., Grześ, M. (eds.) ALA 2011. LNCS (LNAI), vol. 7113, pp. 54–69. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-28499-1_4
Ravichandiran, S.: Hands-on reinforcement learning with python: master reinforcement and deep reinforcement learning using OpenAI gym and TensorFlow. Packt Publishing Ltd. (2018)
Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT Press, Cambridge (2018)
Van Hasselt, H., Guez, A., Silver, D.: Deep reinforcement learning with double q-learning. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 30 (2016)
Viswanathan, V., Lee, C.E., Lees, M.H., Cheong, S.A., Sloot, P.M.: Quantitative comparison between crowd models for evacuation planning and evaluation. Eur. Phys. J. B 87(2), 1–11 (2014)
Wang, Q., Liu, H., Gao, K., Zhang, L.: Improved multi-agent reinforcement learning for path planning-based crowd simulation. IEEE Access 7, 73841–73855 (2019)
Watkins, C.J., Dayan, P.: Q-learning. Mach. Learn. 8(3–4), 279–292 (1992)
Wharton, A.: Simulation and investigation of multi-agent reinforcement learning for building evacuation scenarios. Report, St Catherine’s College, 55 p. (2009)
Xu, D., Huang, X., Mango, J., Li, X., Li, Z.: Simulating multi-exit evacuation using deep reinforcement learning. arXiv preprint arXiv:2007.05783 (2020)
Yang, S., Li, T., Gong, X., Peng, B., Hu, J.: A review on crowd simulation and modeling. Graph. Models 111, 101081 (2020)
Zhang, G., Lu, D., Lv, L., Yu, H., Liu, H.: Knowledge-based crowd motion for the unfamiliar environment. IEEE Access 6, 72581–72593 (2018)
Zheng, S., Liu, H.: Improved multi-agent deep deterministic policy gradient for path planning-based crowd simulation. IEEE Access 7, 147755–147770 (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Abdallah, W., Kanzari, D., Madani, K. (2021). Towards a Deep Reinforcement Approach for Crowd Flow Management. In: Rojas, I., Joya, G., Català , A. (eds) Advances in Computational Intelligence. IWANN 2021. Lecture Notes in Computer Science(), vol 12861. Springer, Cham. https://doi.org/10.1007/978-3-030-85030-2_35
Download citation
DOI: https://doi.org/10.1007/978-3-030-85030-2_35
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-85029-6
Online ISBN: 978-3-030-85030-2
eBook Packages: Computer ScienceComputer Science (R0)