DECISION SUPPORT MODEL FOR POST-EARTHQUAKE EMERGENCY MANAGEMENT BASED ON SEQUENTIAL DATA PROCESSING OF DAMAGE INFORMATION

Abstract

This study presents a method of sequential data processing of seismic damage information for rapid and appropriate decision making in post-earthquake emergency management. First, a real-time estimation of damage rate is conducted using seismic intensity information and fragility relations, providing a prior distribution of damage rate. Next, according to reported damage information, the distribution of damage rate is updated using Bayesian approach in a sequential manner. A technique of SPRT (sequential probability ratio test) is then applied to judge when to make a decision of emergency response. Numerical examples are shown using hypothetical and existing lifeline systems.