Estimating of Analysis Results for Structures with Linear Base Isolation Systems using Artificial Neural Network Model

Abstract

Artificial Intelligence (AI) methods can be used in engineering problem to predict results without the need of long design or optimization processes. In this study, the dynamic analyses of base isolated seismic structures are used in the training of an artificial neural network (ANN) model. The maximum acceleration and displacement of the structure are predicted according to the base isolation period, base isolation damping ratio and pulse period of directivity pulse used in the analyses. The machine learning process was done in three different cases according to number of analysis results used in training. According to the result, the developed ANN model is effective in predicting the behavior of base isolated structures if enough numbers of analyses results were used in machine learning. The ANN model can be used in optimization in future studies.