Seismic vulnerability assessment of masonry buildings using Bayesian belief network (BBN) models

Past earthquakes have caused economic loss, homelessness, and deaths, often due to heavy damage to masonry buildings. Such structures are part of a large existing building stock in Italy designed with obsolete codes or without any seismic provision. Therefore, regional and national strategies to reduce the vulnerability of masonry buildings are necessary. In this study, a Bayesian belief network (BBN) based model is presented for seismic vulnerability assessment of strategic buildings. The proposed model will identify parameters contributing to the building vulnerability and provide a forecast of the damage levels of building due to seismic actions. The building vulnerability assessment accounts for year and quality of construction, as well as the building uses. The conditional probability tables of the BBN model are garnered from the literature, training from existing database and expert knowledge. Existing Italian buildings damaged from recent earthquakes are used for the BBN model calibration and validation.