A Robust Shifting Balance Genetic Algorithm for Topology Optimization of Torsionally Sensitive High-rise Buildings

摘要

Modern tall buildings are getting more and more irregular and complex in shape.During earthquakes, these buildings may vibrate in a lateral torsional manner with significant swaying and twisting, leading to increased deflection, increased structural damage and a higher risk of building collapse.In this paper, a robust evolutionary optimization is presented for seeking the optimal structural topology to best resist lateral-torsional motions of torsionally sensitive buildings under seismic excitations.By incorporating the Sewall Wright’s shifting balance (SB) theory of evolution into the hybrid Optimality Criteria Genetic Algorithm (OCGA), a new hybrid OC-SBGA method is developed to improve the robustness of the evolutionary algorithm in searching the optimal topology and element sizing design of tall asymmetric buildings.A three-dimensional 34-story residential building is used to demonstrate the effectiveness and robustness of the hybrid OC-SBGA method.Results indicate that the proposed method is capable of not only significantly reducing seismic torsional effects in lateral-torsional coupling responses, but also repeatedly attaining the same optimal topology design among different independent optimization runs.