A Two-Stage Genetic Algorithm for Beam–Slab Structure Optimization

Beam–slab structures account for 50–65% of a building’s total dead load and contribute to 20% of the overall cost and CO2 emissions. Despite their importance, conventional beam–slab structural optimization methods often lack search efficiency and accuracy, making them less effective for practical en...

Full description

Saved in:
Bibliographic Details
Published inBuildings (Basel) Vol. 14; no. 9; p. 2932
Main Authors Yang, Zhexi, Lu, Wei-Zhen
Format Journal Article
LanguageEnglish
Published Basel MDPI AG 01.09.2024
Subjects
Adaptability
Algorithms
Analysis
Artificial intelligence
beam–slab structure
Carbon dioxide
Carbon dioxide emissions
Case studies
component optimization
Dead loads
Design
Efficiency
Engineers
genetic algorithm (GA)
Genetic algorithms
Genetic research
Heuristic methods
layout optimization
Load
Load transfer
Machine learning
Neural networks
Optimization
Public housing
Solution space
Static loads
Statistical analysis
topology optimization
Variables
Hong Kong
China
Hong Kong China
Online AccessGet full text

Cover

More Information
Summary:Beam–slab structures account for 50–65% of a building’s total dead load and contribute to 20% of the overall cost and CO2 emissions. Despite their importance, conventional beam–slab structural optimization methods often lack search efficiency and accuracy, making them less effective for practical engineering applications. Such limitations arise from the optimization problem involving a complex solution space, particularly when considering components’ arrangement, dimensions, and load transfer paths simultaneously. To address the research gap, this study proposes a novel two-stage genetic algorithm, optimizing beam–slab layout in the first stage and component topological relationships and dimensions in the second stage. Numerical experiments on the prototype case indicate that the algorithm can generate results that meet engineering accuracy requirements within 100 iterations, outperforming comparable algorithms in both efficiency and accuracy. Additionally, this heuristic approach stands out for its independence from prior dataset training and its minimal parameter adjustment requirement, making it highly accessible to engineers without programming expertise. Statistical analysis of the algorithm’s optimization process and case studies demonstrate its robustness and adaptability to various beam–slab structural optimization problems, revealing its significant potential for practical engineering scenarios.
ISSN:2075-5309
2075-5309
DOI:10.3390/buildings14092932