Computational Tool for Post-Earthquake Evaluation of Damage in Buildings

Abstract

A method and a computational tool oriented to assist the damage and safety evaluation of buildings after strong earthquakes is described in this article. The input of the model is the subjective and incomplete information on the building state, obtained by inspectors which are possibly not expert professionals of the field of building safety. The damage levels of the structural components are usually described by linguistic qualifications which can be adequately processed by computational intelligence techniques based on neuro-fuzzy systems what facilitate the complex and urgent tasks of engineering decision-making on the building occupancy after a seismic disaster. The hybrid neuro-fuzzy system used is based on a special three-layer feedforward artificial neural network and fuzzy rule bases and is an effective tool during the emergency response phase providing decisions about safety, habitability, and reparability of the buildings. Examples of application of the computer program are given for two different building classes.