Comparison of meta-heuristic approaches for the optimization of non-linear base-isolation systems considering the influence of superstructure flexibility

Abstract

The main aim of this study is to investigate the meta-heuristic approaches called Harmony Search (HS), Grey Wolf Optimizer (GWO), Teaching-Learning Based Optimization (TLBO), and Jaya Algorithm (JA) in the optimization process of non-linear base isolation systems under near-fault earthquakes considering the effect of the superstructure flexibility. The optimization processes were performed by achieving the objective function set as minimizing the peak top floor acceleration to peak ground acceleration ratio with and without base displacement limits. The context of the optimization process, the analytical model of the superstructure, and the number of isolators are considered as the design constants. The mechanical parameters of the non-linear isolation systems, such as the isolation system period, the total characteristic strength ratio, and the yield displacement, are determined as the independent design variables. According to the results obtained in this study, GWO, JA, and TLBO algorithms can be more suitable for solving such design problems among the considered algorithms. Although the objective function values are amplified with the increase in the superstructure flexibility, these values remain around 1.0, even for the most stringent base displacement limit. It can be expressed as a successful seismic performance, especially considering such strong near-fault ground motions, which are the most challenging types for seismically isolated buildings.