Neural networks for the rapid seismic assessment of existing moment-frame RC buildings

The study presented in this paper analyses and investigates the possibility of introducing a general rapid methodology based on an artificial neural network (ANN) to assess seismic response existing reinforced concrete (RC) buildings. Starting from investigations carried out buildings located outskirts Bologna, 928 finite element models have been developed basis most recurrent data. input parameters representing characteristics structures were systematically varied and, through modal dynamic non-linear static analyses, outputs recorded. resulting dataset was used create function, ANN, that can reliably predict behaviour RC structure. Finally, by means k-fold cross-validation, instruction function optimised simultaneously verified, obtaining coefficient determination for totality samples previously unseen cases 0,94 0,88, respectively. result obtained not only aims at enriching framework subject, increasing awareness issues affecting building typology, but also constitutes prioritization system could highlight need structural renovation.