Modeling the Evacuation of Large-Scale Crowded Pedestrian Facilities

Most of these accidents occur as pedestrians try to evacuate the facility in panic mode. Thus, increasing attention has been given to planning for the evacuation of these facilities by considering of all possible emergencies. The objective is to minimize the period required to evacuate a facility while ensuring the safety of all pedestrians. Evaluation of the effectiveness of an evacuation plan in large facilities through actual field testing would be a challenging task because of the complexity of the logistics and high cost required to enact realistic scenarios that involve large numbers of pedestrians. Nonetheless, pedestrian modeling techniques, including microsimulation, have emerged as a valid alternative. Considerable research in the area of pedestrian modeling and crowd dynamics has been reported in the literature. For example, Okazaki was among the first to present a microsimulation model for pedestrian movement (1). The model adopts the concept of magnetic force to emulate how pedestrians avoid obstacles during their movements. Gipps and Marksjo present an approach in which the movement of pedestrians to a target location is modeled as a function of the trade-off between the cost and the benefit associated with each potential movement (2). A social force model that builds on the ideas of kinetic theories of traffic flow was developed by Helbing and Molnar (3). Blue and Adler introduced the cellular automata (CA) technique to represent pedestrian movements (4). The area is divided into regular cells such that each cell can be occupied by one person at a time. Still developed a model (the Legion model) that simulates the crowd as an emergent phenomenon using simulated annealing and mobile CA (5). That model is based on the interaction of four parameters: objective, motility, constraint, and assimilation. Abdelghany et al. extended the CA approach through implementing a simulation-assignment framework that allows each individual to choose and update her or his path in the facility as a function of the evolving congestion (6). Examples of the use of pedestrian microsimulation in evaluating pedestrian evacuation can be found in the work of Watts (7 ), Lovas (8), Thompson and Marchant (9), Helbing et al. (10), Santos and Aguirre (11), and Parisi and Dorso (12). These studies vary by the size of the modeled facilities and their geometric complexity, as well as the modeling approach used and the assumptions that capture pedestrian behavior. These studies also vary in how realistically they represent the evacuation scenarios modeled. Achieving a realistic representation of an evacuation scenario requires a modeling framework that captures the different behavioral rules that govern the dynamics of pedestrian movements. These rules describe how each pedestrian chooses, for example, her or his exit gate, the path to this gate, and the walking speed along the chosen path. The rules should also be able to capture the frequency of updating of all these decisions as a function of the evolving congestion in Modeling the Evacuation of Large-Scale Crowded Pedestrian Facilities