arXiv · 2611.22502
Crowd model with anticipation and interactions: an exclusion process based mean-field game
Abstract
Mean-field games (MFG) is a powerful framework to describe competitive optimization processes. A natural field of application is crowd dynamics, where pedestrians are considered to optimize their path, while interacting with the others. Despite its success to incorporate anticipation in describing pedestrian behavior, the mean-field games framework neglects microscopic granular effects, rendering its results less realistic in various applications (bottlenecks, evacuations etc.). In this work, we consider an explicitly interacting microscopic model, namely an asymmetric simple exclusion process and derive the corresponding mean-field game. This mean-field game is applied to an evacuation setting and the differences from the standard MFG based on non-interacting Langevin dynamics, as well as the correlations induced by the granular interactions are discussed.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Konstantinos Koutsomitis, Cécile Appert-Rolland, Denis Ullmo. 2026-11-17. Crowd model with anticipation and interactions: an exclusion process based mean-field game. https://arxiv.science/abs/2611.22502
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.