Optimization of Evacuation Route considering Uncertainty of Undetected Fires in Real-Time Evacuation Guidance under Post-Earthquake Fires

Abstract

After an earthquake, real-time evacuation guidance, which guides inhabitants to safe site from danger area based on fire information, is important as an emergency measure against simultaneous fires. In this study, we optimized a realtime route considering incompleteness of fire information. First, we constructed the stochastic model called “evacuation probability”, which is the probability that evacuee can reach evacuation site without encountering an undetected fire. Second, we maximize an evacuation probability on a real-time route under condition of securing a specified value of “safe margin of evacuation”, which means maximum tolerance on a deviation between a forecasted fire speed and an actual one. Finally, we discussed characteristics of the route with visualizing and analyzing results of simulations.