Impact of Cultural Differences on Crowd Dynamics in Pedestrian and Evacuation Domains

摘要

This report results from a contract tasking Bar Ilan University as follows: Accounting for Culture in Agent-Based Pedestrian Crowd Simulation. Accurate models of crowd dynamics are an important challenge in multi-agent systems and agent-based social simulation. Crowd models are able to predict the resulting macro level behavior from micro level interactions. However, many existing crowd models do not yet account for cultural factors in crowd behavior, and even more so, for crowds composed of members of different cultures. In this paper we examine the impact of cultural differences on the crowd dynamics in pedestrian and evacuation domains. In the pedestrian domain we relate to recorded pedestrian data in five different countries: Iraq, Israel, England, Canada and France and characterize these cultures based on cultural attributes at the individual level: personal spaces, speed, avoidance side and group formations. We use an agent-based simulation to investigate the impact on the resulting macro level behavior, such as pedestrian flow, number of collisions, etc. We also examine the impact of mixed-culture pedestrians on the resulting macro-level behavior. We quantitatively validate the simulation against data from movies of human crowds, in different countries. In the evacuation domain, we use an established simulation system to investigate cultural differences reported in the literature, and additionally explore the resulting macro level behavior.