Linear Dynamic Responses under Nonstationary Forces with Artificial Neural Network Model

Abstract

The objective of this paper is to propose the artificial neural network (ANN) model as a surrogate model for determining the dynamic responses of shear frame structure under nonstationary excitation due to the seismic forces. A multi-layer feed-forward neural network is employed particularly in ANN architecture for mapping a large number of generated non-stationary excitations that based on the universal approximation theorem as input to output of stochastic dynamic responses analysis. The proposed idea and computational procedure is illustrated through a time-domain analysis of the simplified 3-DOF linear dynamic system subjected to non-stationary ground motion excitation. From the obtained numerical results shown the potential of the proposed methodology in terms of computational time and accuracy.