Evaluation of seismic response of buildings and probabilistic damage estimation applying bayesian model updating

Abstract

A method for evaluating seismic response of each story of a building utilizing signals of sensors installed on limited floors of the building is proposed, where the design model of the building is updated to a model closer to reality by applying Bayes' theorem. A fully probabilistic framework for estimation of damage probability of each component on each floor is also developed. The proposed method is applied to seismic response analysis results, which shows extremely precise evaluation ability of the method. Visualization of damage probability is also considered to be effective for decision making after earthquakes.