Multimodal size, shape, and topology optimisation of truss structures using the Firefly algorithm

Authors: Leandro Fleck Fadel Miguel, Rafael Holdorf Lopez, Letícia Fleck Fadel Miguel

Abstract: 
This paper presents an efficient single-stage Firefly-based algorithm (FA) to simultaneously optimise the size, shape and topology of truss structures. The optimisation problem uses the minimisation of structural weight as its objective function and imposes displacement, stress and kinematic stability constraints. Unstable and singular topologies are disregarded as possible solutions by checking the positive definiteness of the stiffness matrix. Because cross-sectional areas are usually defined by discrete values in practice due to manufacturing limitations, the optimisation algorithm must assess a mixed-variable optimisation problem that includes both discrete and continuous variables at the same time. The effectiveness of the FA at solving this type of optimisation problem is demonstrated with benchmark problems, the results for which are better than those reported in the literature and obtained with lower computational costs, emphasising the capabilities of the proposed methodology. In addition, the procedure is capable of providing multiple optima and near-optimal solutions in each run, providing a set of possible designs at the end of the optimisation process.

Keywords:
Truss size
Shape and topology optimisation
Firefly algorithm
Multimodal optimisation
Mixed continuous-discrete variable problems
Structural optimisation Single-stage procedure

Published in: Advances Engineering Software (Volume 56, February 2013)

Publisher: Elsevier

ISSN Information: 0965-9978

Multimodal size, shape, and topology optimisation of truss structures using the Firefly algorithm

Bình luận của bạn
*
*
*
*
 Captcha

Logo Bottom

Địa chỉ: 268 Lý Thường Kiệt, P.14, Q.10, TP.HCM           Tel: 38647256 ext. 5419, 5420           Email: thuvien@hcmut.edu.vn

© Copyright 2018 Thư viện Đại học Bách khoa Tp.Hồ Chí Minh 

Thiết kế website Webso.vn