Publication detail

Multilevel Optimization of Complex Composite Structures with Variable Stiffness, Part I: Description of the Optimization Algorithm

SYMONOV, V.

English title

Multilevel Optimization of Complex Composite Structures with Variable Stiffness, Part I: Description of the Optimization Algorithm

Type

journal article in Web of Science

Language

en

Original abstract

A multilevel hybrid optimization algorithm is proposed for the weight minimization of complex composite structures with variable stiffness. The key improvements of the algorithm in comparison with the previous ones are the simplification of laminate definition at the upper level using only three parameters, an original guide-based genetic algorithm, a one-dimensional search algorithm based on the "Golden section" method and adopted to integer variable, a parallel optimization, and a simplified upper-to-lower-level coordination procedure.

English abstract

A multilevel hybrid optimization algorithm is proposed for the weight minimization of complex composite structures with variable stiffness. The key improvements of the algorithm in comparison with the previous ones are the simplification of laminate definition at the upper level using only three parameters, an original guide-based genetic algorithm, a one-dimensional search algorithm based on the "Golden section" method and adopted to integer variable, a parallel optimization, and a simplified upper-to-lower-level coordination procedure.

Keywords in English

two-level optimization algorithm; composite structure; finite-element analysis; genetic algorithm

Released

14.03.2023

Publisher

Springer

ISSN

0191-5665

Volume

59

Number

1

Pages from–to

65–76

Pages count

12

BIBTEX


@article{BUT183239,
  author="Volodymyr {Symonov},
  title="Multilevel Optimization of Complex Composite Structures with Variable Stiffness, Part I: Description of the Optimization Algorithm",
  year="2023",
  volume="59",
  number="1",
  month="March",
  pages="65--76",
  publisher="Springer",
  issn="0191-5665"
}