Author(s): Hibiki Motoi; Yukikazu Yamaguchi

Linked Author(s): HIBIKI Motoi

Keywords: Tsunami; Multi Agent Simulation; Simulation; Evacuation; Disaster

Abstract: Underground shopping mall, which have become increasingly large and complex in recent years, have become important walking spaces for an unspecified number of users. Therefore, underground shopping mall need to ensure appropriate urban functions and at the same time promote disaster prevention and mitigation measures against large-scale earthquakes and accompanying tsunamis. In order to evacuate a large number of visitors to a safe place in the event of a disaster, it is effective to predict evacuation behavior in advance and improve evacuation problems. However, it is not realistic to conduct an experiment in a large-scale underground shopping mall where visitors are actually asked to take evacuation actions in a disaster situation. For this reason, multi-agent systems are considered to be an effective method. In this study, we develop a multi-agent simulation model that takes into account the behavioral characteristics of individual visitors and their interactions in the complex space of an underground shopping mall. To construct the multi-agent simulation model, we use data obtained from a preliminary survey. In the preliminary survey, the number of visitors in the target area, their walking speed, walking speed on the stairs, and which exit they evacuate to were investigated. Then, the evacuation of the visitors is reproduced in the simulation, and the evacuation guidance is executed. The method of evacuation guidance is to guide the evacuees at the intersection of the underground shopping mall and guide them to the nearest exit where the number of people staying in front of the exit is small. By doing so, we believe that we can prevent the concentration of visitors at a particular exit and reduce the occurrence of congestion in front of the exit by dispersing the visitors, thereby shortening the evacuation completion time. The purpose of this study is to clarify the effect on crowd evacuation by changing the timing of the start of evacuation guidance. Specifically, we will conduct evacuation simulations by guiding visitors in the target area at different times, and analyze the optimal timing to see how much it affects the evacuation completion time and the crowded evacuation exits. In addition, the number of visitors to the city will be varied, and the optimal timing to start evacuation guidance for each different number of visitors will be analyzed. By conducting this research, it will be possible to provide information that will contribute to the optimization of evacuation guidance in the event of a disaster.

DOI: https://doi.org/10.3850/IAHR-39WC2521711920221819

Year: 2022