A multi‐objective structural optimization method for serviceability design of tall buildings

TL;DR: A multi-objective optimization method for serviceability design of tall buildings in multi-hazard environments. The method incorporates seismic and wind effects and utilizes a neural network-based surrogate model for efficient optimization.

Abstract: Structural optimization design aims to identify optimal design variables corresponding to a minimum objective function with constraints on performance requirements. To this end, many optimization frameworks have been proposed to determine optimal structural systems that are subjected to seismic and wind hazards in isolation. However, some modern tall buildings are sensitive to seismic and wind excitation owing to their complex structural systems and geographic regions. Therefore, a proper structural optimization method for such buildings is required to ensure that the expected performance is achieved in a multi‐hazard scenario. This study proposes a multi‐objective serviceability design optimization methodology for buildings in multi‐hazard seismic and wind environments by combining optimality criteria and the nondominated sorting genetic algorithm II (NSGA‐II). Seismic and wind effects can be instantaneously updated due to changes in the structural dynamic properties during the optimal design process. A neural‐network‐based surrogate model with self‐updating is proposed to predict the structural natural frequency so that the overall computation time of the optimization process can be reduced. The proposed method was used to optimize a 50‐story frame‐tube building and was compared against the general genetic algorithm and general NSGA‐II to verify the feasibility and effectiveness.