A Hybrid macro-micro pedestrians evacuation model to speed up simulation in road networks

Ngoc-Anh T. Nguyen (1) , Jean-Daniel Zucker (2) , Huu-Du Nguyen (3) , Alexis Drogoul (1) , Duc An Vo (1)

1 MSI - Modélisation et Simulation Informatique de systèmes complexes
2 UMMISCO - Unité de modélisation mathématique et informatique des systèmes complexes [Bondy]
3 University of Science

The major methodologies of crowd simulation in a dynamic environments are either based on micro or macro models. These two types of models represent the trade-off between the level of detail and efficiency. The domain of pedestrian flow simulation on road networks is no exception and theories rely either on equation based model or agent based models. There is a growing interest in hybrid modeling that com- bines both of these types. This paper addresses the problem of combining both micro and macro models of pedestrians movement to speedup sim- ulations. Hybrid model uses efficient macro modeling in part of the road networks that do not require a fine grained model and more detailed but less efficient micro modeling in critical locations. One key issue raised by such an approach and discussed is the consistency of the resulting hybrid model. Preliminary results presented in this article is a proof of concept that the use of hybrid model to simulate evacuation plan in road networks may be more efficient than the use of micro model alone.

Mots clés: Crowd movement; Hybrid modeling; road networks; multi agent systems

Domaines: Modélisation et simulation

Ngoc-Anh T. Nguyen, Jean-Daniel Zucker, Huu-Du Nguyen, Alexis Drogoul, Duc An Vo. A Hybrid macro-micro pedestrians evacuation model to speed up simulation in road networks. AAMAS, May 2011, Taipei, Taiwan. ⟨10.1007/978-3-642-27216-5_28⟩. ⟨hal-00736474⟩