A micro-macro approach to modeling the flocking of crowd dynamics
Date:
Liang, H.*, Ye, Y., & Sun, J. (2024, January 7-11). A micro-macro approach to modeling the flocking of crowd dynamics [Poster Presentation]. The 103rd Annual Meeting of Transportation Research Board, Washington DC, United States.
Abstract: Understanding and predicting the flocking crowd dynamics is crucial for crowd management and disaster preparedness. It is widely recognized that pedestrians not only act based on their individual preferences but also respond to global information from the crowd and geometry. While existing model-based studies fall short in capturing the unique behaviors observed in dense crowds, characterized by high levels of aggregated pressure and panic sentiments, this study proposes a novel micro-macro approach to model these flocking phenomena. At the microscopic level, pedestrians adjust their speed and positions dynamically according to a modified version of the Social Force Model. At the macro level, a T-Gaussian distance-dependent measure is developed to determine the macroscopic variables. Additionally, a cost potential model and a crowd pressure model are introduced to describe the collective behavior of the crowd. Through designed numerical tests, the proposed model demonstrates its ability to reproduce observed flocking phenomena, including fan-shaped congestion, aggregated crowd pressure, and the occurrence of a second peak on the fundamental diagram. This model framework has the potential to be applied in the predictive management of dense crowds, allowing for better crowd control and planning strategies.