Teaching-Learning Based Optimization of Nonlinear Isolation Systems under Far Fault Earthquakes

Seismic isolation systems exposed to far-fault earthquakes can reduce floor accelerations and story drift ratios to acceptable levels. However, they exhibit different structural performances in each earthquake due to different excitation frequency contents. By optimizing the isolation system parameters, their performance may be maintained at the best level under different far-fault earthquakes. In this study, the optimization of the parameters of the nonlinear isolation system of a 5-story benchmark building is performed by Teaching-Learning Based Optimization (TLBO) algorithm to minimize peak floor accelerations under historical far-fault earthquakes with and without exceeding a specified base displacement limit. According to the results of the analyses, it can be said that TLBO algorithm is a robust algorithm with low standard deviations for determining optimum nonlinear isolation system parameters.