Utilization of artificial intelligence for estimation of fundamental time period of plan asymmetric buildings

Estimation of the fundamental time period forms a critical step in the seismic design process. The design force in a building is inversely proportional to the fundamental time period. Therefore, underestimation of fundamental time period leads to overestimation of the design forces which results in uneconomical design and vice versa. The expressions proposed by the seismic design codes have ignored the aspect of structural irregularity in the buildings which leads to inaccurate estimation of the seismic design forces and inappropriate design. This is one of the major reasons for the failure of irregular buildings during previous earthquakes. This research work focusses on addressing the shortcomings of the seismic design codes. To achieve this purpose 200 building frames with different magnitude of plan irregularity have been modeled and subjected to the dead and live load. To train the input and output data feed forward Artificial neural network based on Levenberg - Marquardt algorithm is used. Based on the analysis results a seismic response databank has been prepared using E-Tabs 2009 software. Then, regression analysis has been conducted on this database to propose new equations to estimate the fundamental time period for plan irregular buildings. The proposed equations have been compared with the code approaches to demonstrate the efficiency of the proposed equations.