Full Length Research Paper
Abstract
The design of dome structures is optimized by a genetic algorithm methodology with multi populations. In order to increase the convergence degrees of optimal designs obtained, exploitation and exploration capacities of the genetic algorithm methodology are enhanced. In this regard, a radial basis neural network and a new design strategy based on provisions of LRFD_AISC V3 specification are implemented into the optimization procedure. Furthermore, the size-shape-topology design variables are simultaneously utilized to automatically generate both sphere and ellipse-shaped dome structures. The computational performance of the proposed optimization approach named as enhanced genetic algorithm with multiple populations (EGAwMP) is evaluated considering the design optimization of two dome structures. It is demonstrated that the proposed optimization approach succeed in obtaining optimal designs with higher converged degree. Furthermore, it is displayed that using the size-shape-topology design variables for a sphere-shaped dome structure increases the optimality quality of designs compared to those obtained by using size-shaped-design variables for an ellipse-shaped dome structure. Consequently, the proposed optimization approach is recommended to optimize the design of dome structures as an intelligence optimization tool.
Key words: Genetic algorithm, multiple populations, neural network, domes, LRFD-AISC.
Copyright © 2024 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0