MODELING IN SOCIAL SYNAMICS: A DIFFERENTIAL APPROACH
This research proposes a major expansion of a simulation of pedestrian motion based on the model of pedestrian motion by Helbing-Moln ar-Farkas-Vicsek (HMFV). This model uses a combination of social and physical forces to determine how the individuals in the model move and interact. The purpose of this expansion is to fold characteristics concerning age, gender, ethnicity, and culture differences into the model, so as to be able to capture reliable differences due to these four factors. Such results could then be used to improve our ability to model social dynamics. This will be done by giving each individual in the simulation, certain traits and characteristics, related to culture, ethnicity, gender, and age. The simulations would then demonstrate how these variations in individual characteristics can affect individual behavior and group behavior, as in a crowd. These four factors (culture, ethnicity, gender, and age) will certainly provide a sufficient space to be interesting. However, there are also interesting cases wherein one would want to model groups of individuals with similar characteristics. These shall be concentrated on at first, expanding the dimension of the space as experience is gained. The distribution of these traits in the simulated populations will be stochastic so as to have flexibility to model a variety of situations (e.g. night club fire, a sporting event, a subway platform). Having stochastic distributions of traits will allow the opportunity to determine how sensitive the overall crowd behavior is to small changes in distributions of the traits. It will also allow us to model "worst case" scenarios so that those engaged in crowd control (or building design) can prepare for the most detrimental outcome.
Recent world events prompt research in understanding crowd behavior in emergencies such as nightclub fires, stadium accidents, and subway bombings. The continual threat of global terror and the potential for mass panic could lead to more fatalities of people who are crushed or trampled. This project will use mathematical modeling and simulations to construct models of pedestrian interactions under various situations. The results will be validated using actual crowd movement data obtained from videos from various scenarios. The findings will provide understanding of the influence and importance of social variability under various conditions. The goal of this effort is to provide a more realistic simulation model of crowd behavior under both panic and non-panic conditions by considering individual differences in age, gender, culture and ethnicity. This would be a valuable tool for police departments, the military, and others to use for event handling or in terrorist situations. It could also be used for building design to accommodate cultural differences.
