Predicting the peak structural displacement preventing pounding of buildings during earthquakes

The aim of the present paper is to verify the effectiveness of the artificial neural network (ANN) in predicting the peak lateral displacement of multi-story building during earthquakes, based on the peak ground acceleration (PGA) and building parameters. For the purpose of the study, the lumped-mass multi-degree-of-freedom structural model and different earthquake records have been considered. Firstly, values of stories mass and stories stiffness have been selected and building vibration period has been automatically calculated. The ANN algorithm has been used to determine the limitation of the peak lateral displacement of the multi-story building with different properties (height of stories, number of stories, mass of stories, stiffness of stories and building vibration period) exposed to earthquakes with various PGA. Then, the investigation has been focused on critical distance between two adjacent buildings so as to prevent their pounding during earthquakes. The proposed ANN has logically predicted the limitation of the peak lateral displacement for the five-story building with different properties. The results of the study clearly indicate that the algorithm is also capable to properly predict the peak lateral dis-placements for two buildings so as to prevent their pounding under different earthquakes. Subsequently, calculation of critical distance can also be optimized to save the land and provide the safety space between two adjacent buildings prone to seismic excitations.