Predicting the Influence of Soil–Structure Interaction on Seismic Responses of Reinforced Concrete Frame Buildings Using Convolutional Neural Network

Most regional seismic damage assessment (RSDA) methods are based on the rigid-base assumption to ensure evaluating efficiency, while these practices introduce factual errors due neglecting soil–structure interaction (SSI). Predicting influence of SSI responses regionwide structure portfolios remains a challenging undertaking, as it requires developing numerous high-fidelity, integrated models capture dynamic interplay and uncertainties in structures, foundations, supporting soils. This study develops one-dimensional convolutional neural network (1D-CNN) model efficiently predict what degree considering would change inter-story drifts base shear forces RC frame buildings. An experimentally validated finite element is developed simulate nonlinear behavior building-foundation–soil system. Subsequently, database comprising input data (i.e., structural soil parameters, ground motions) output predictors changes story drift shear) constructed by simulating 1380 pairs fixed-base versus soil-supported structures under earthquake loading. large-scale dataset used train, test, identify optimal hyperparameters for 1D-CNN quantify demand differences shears SSI. Results indicate has superior performance, absolute prediction coefficients maximum within 9.3% 11.7% 80% cases testing set. The deep learning can be conveniently applied enhance accuracy RSDA buildings updating their where no considered.