Cost optimization of structures using a genetic algorithm with Eugenic Evolutionary theory

This paper illustrates the cost optimization of steel frame structures by mean of genetic algorithm developed from the Eugenics Evolutionary theory. The aim is to obtain a final structure with a minimum cost. To this end, a modified multiple objective function has been defined. This considers cost a...

Full description

Saved in:
Bibliographic Details
Published in:Structural and multidisciplinary optimization 2016-08, Vol.54 (2), p.199-213
Main Authors: Prendes-Gero, María-Belén, Álvarez-Fernández, Martina-Inmaculada, López-Gayarre, Fernando, Drouet, Jean-Marc, Junco, Julio Rodríguez-Vigil
Format: Article
Language:English
Subjects:
Computational Mathematics and Numerical Analysis
Engineering
Engineering Design
Evolutionary algorithms
Frame structures
Genetic algorithms
Genetic engineering
Minimum cost
Multiple objective analysis
Optimization
Parameters
Research Paper
Steel frames
Steel structures
Structural members
Theoretical and Applied Mechanics
Welded joints
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper illustrates the cost optimization of steel frame structures by mean of genetic algorithm developed from the Eugenics Evolutionary theory. The aim is to obtain a final structure with a minimum cost. To this end, a modified multiple objective function has been defined. This considers cost as the result of a summary where elements like welds, simple connections, or the number of structural elements, have an influence on the final result. According to the Eugenics Evolutionary theory, a new selection operator has been developed in a way that leads to all members of the population being able to have descendants and avoids the loss of any kind of genetic material. In addition, the penalization coefficients have been optimised and the effect of parameter setting has been investigated, to achieve convergence faster through penalising the most expensive structures and looking for the optimum range of parameters’ value. The result is a robust genetic algorithm which, compared with others, achieves better optimum individuals and does not stop at local minima. Finally, two different two-dimensional truss frames have been optimized and the results have been compared with those obtained using different methods of selection like elitism, steady-state replacement, roulette wheel, and tournament selection.
ISSN:1615-147X
1615-1488
DOI:10.1007/s00158-015-1249-5