.png){kind=icon}
.png){kind=icon}
Return
Application of machine learning methods for intelligent optimization in structural engineering
.png){kind=icon}
PRE
AI
DOI：10.1080/0305215X.2025.2592039.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Conventional structural design is a labour-intensive and inefficient task, requiring a lot of manual modelling and parameter tuning. Intelligent structural optimization design is limited to simple scenarios, creating barriers in complex systems. This study aims to solve the optimization problem in structural engineering based on machine learning methods and optimization algorithms. A multilateral force-resisting system is adopted for analysis, considering structural constraints according to standard codes, with its total cost to be minimized. Surrogate models are established for structural performance prediction, using the grid search and mechanical analysis-based feature engineering method. The structural intelligent optimization is completed by the surrogate-based optimization (SBO) method, with the algorithm performances compared. The results show that the optimized structure has great lateral force-resisting performance, complying with safety demands and economic benefits. SBO achieves high efficiency and robustness in structural intelligent optimization. This framework has general significance and could be extended to various engineering projects.
Keywords:
Structural intelligent optimization
machine learning algorithm
metaheuristic algorithm
surrogate model
multilateral force-resisting system
Journal
IF:
2.2
Papers:
105
Citations:
3.8K