Multiple exits evacuation algorithm for real-time evacuation guidance

Abstract

Most studies for minimizing total evacuation time do not take into account aspects of realistic evacuation guidance because they focus on minimizing evacuation time arithmetically. In the mentioned study, occupants in one space can be divided and move to different exits in order to minimize the evacuation time. However, in an emergency situation, it is practically difficult to guide occupants in a space to different directions, and may confuse them significantly. For this reason, this study proposed a multiple exits evacuation algorithm (MEEA) that guide the occupants in one space to the same exit. The MEEA is based on graph theory and computes a process of exits assignment of the nodes, leading to the division of the spaces based on the exits. Each exit competitively absorbs nodes, repeating until evacuation times of exits are balanced and the total evacuation time is minimized. In order to verify MEEA, this study used evacuation simulators based on cellular automata called EgresSIM to compare the evacuation results of well-known evacuation models EVACNET4 and MEEA.