被害推定の逐次更新機能を有する緊急対応の意思決定支援シミュレータの開発

概要

論文の詳細を見る
In post-earthquake emergency, decision makers are often involved in a trade-off between rapidness and preciseness of important actions. To best balance these two requisites, a decision support simlulator has been developed on the basis of two different sources of damage information: real-time damage assessment and actual damage information. Damage estimates are updated in a sequential manner following Bayesian updating theory. SPRT(Sequential probability ratio test)-based algorithm provides a rational basis for adequate action policies. Numerical examples demonstrate the effective procedures of the implemented algorithms. Total damage can be outlined in terms of probability distributions and their central values. The hardest-hit areas in need of prioritized relief can be identified.
地域安全学会の論文