Guide optimization in pedestrian emergency evacuation

An extended guided crowd dynamics model is proposed to investigate the guide optimization during emergency evacuation. The initial assignment scheme for guides based on clustering algorithm, taking the numbers of both informed and uninformed pedestrians into consideration, is studied. The guide choice method for informed followers based on an exponent model, considering the distance to the guide and the follower quantity, is further investigated. On the basis of the modeling method of this paper, evacuation behavior dynamics under guides is explored, from which the leading role of guides can be clearly observed. The effects of guide quantity, parameter in the guide choice method and the size of visual field on evacuations are analyzed. Simulation results indicate that the evacuation efficiency gradually increases with the increase of guide quantity until it reaches a certain level, and the optimal number of guides is closely related to the initial distribution of pedestrians. The combination of density factor and distance factor with a reasonable proportion when determining which guide is very necessary. Pedestrians have stronger evacuation ability within a certain period of time when continuously increasing their visual radii. The study can provide theoretical suggestions for guiding pedestrian evacuation under emergencies. (C) 2019 Elsevier Inc. All rights reserved.