This repository contains the dataset and source code developed for the study:
“Reinforced Concrete Deep Beams Optimization Using Data-Driven Parametric Performance Indicators.”
DOI: https://doi.org/10.54878/xqyna996
The project presents a computational multi-objective optimization framework for reinforced concrete deep beams. An empirical dataset is processed through data normalization, parametric search, and multi-objective evaluation within a bounded evolutionary domain. Four parametric performance indicators—Cross-sectional Efficiency, Reinforcement Contribution Ratio, Tensile Mechanism, and Structural Gradient—are used to evaluate structural performance under a consistent unit system (mm, kN, MPa). The Octopus MOO Optimizer is employed to obtain a stabilized Pareto front, with post-processing using Python-based trendline regression to refine and interpret optimal design outcomes for practical engineering applications.
Cite the Models, and Source Code:
- Experimental dataset and analysis
- Implementation of four parametric performance indicators
- Multi-objective optimization framework using Octopus MOO
- Python post-processing with trendline regression
This repository accompanies the published article and is intended to support reproducible research and the application of machine learning techniques in reinforced concrete structural engineering.