Mengchun Xie, Mitsutoshi Murata, Yusuke Muraki
tsunami evacuation, evacuation guidance simulation, multi-agent systems, OpenStreetMap, instruction agent, evacuee agent, disaster prevention plan
In order to reduce human casualties in the aftermath of earthquake-related tsunamis, it is vital that proper and efficient evacuation plans be set in place and practiced before such disasters occur. Evacuation simulations can provide effective tools for enhancing disaster prevention in areas and regions where it is difficult to carry out actual disaster drills or evacuation-related experiments, but there are few evacuation-related guidance simulations that can be used to examine the influence of placement conditions and evacuation leader behaviors on the evacuation rate. However, if simulating optimum placement conditions and the evacuation leader behavior was to become possible, such simulations would be useful for examining municipal or regional disaster prevention plans. In this paper, we report on the construction of a multi-agent evacuation guidance simulation consisting of instruction agents and evacuee agents, and experiments in which it was used to investigate the effect of evacuation guidance on the different conditions related to a post-earthquake tsunami evacuation. In addition, we also simulated the current situation in Gobo City, Wakayama Prefecture, Japan and confirmed that this model could be successfully applied to that city. From the evaluation guidance experiments, it was revealed that the evacuation rate is lower in situations where there are insufficient numbers of evacuation leaders, and when waiting for an excessive time to collect additional numbers of evacuees.
Cite this paper
Mengchun Xie, Mitsutoshi Murata, Yusuke Muraki. (2017) Tsunami Evacuation Guidance Simulation using Multi-agent Systems Based on OpenStreetMap. International Journal of Environmental Science, 2, 231-237