Buy article PDF
The purchased file will be sent to you
via email after the payment is completed.
US$ 35
|
Computers and Concrete Volume 35, Number 1, January 2025 , pages 97-111 DOI: https://doi.org/10.12989/cac.2025.35.1.097 |
|
|
|
Predicting deterioration components of reinforced concrete columns based on stacking ensemble learning model |
||
A.Khoshkroodi, H.Parvini Sani and M.Sadeghi
|
||
| Abstract | ||
| Deterioration components (DCs) of reinforced concrete columns are important for the seismic performance assessment of reinforced concrete (RC) structures. DCs in RC columns includes: plastic chord rotation from yield to cap (Ɵp), post-capping plastic rotation capacity from the cap to point of zero strength (Ɵpc), and normalized energy dissipation capacity (lambda). This paper investigates several machine learning (ML) algorithms for the prediction of DCs, referred to as ML-DCs, based on the results of 255 experimental tests conducted on reinforced concrete columns from 1973 to 2002. Also in this research, for the prediction of DCs, a learning-based prediction model is developed in a stacking ensemble framework, that is the training of several different basic models and their combination through the training of a meta-model to make a final prediction based on predictions made by the basic models. AdaBoost, Random Forest (RF), Support Vector Regression (SVR) and XGBoost algorithms are base learners. Among the ML algorithms, stacking ensembles showed the best results with a correlation coefficient (R2) of 0.778 for Ɵp, 0.703 for Ɵpc, and 0.809 for lambda. The results of ML algorithms indicate that the ML models are more effective than the empirical relationships that are based on the experimental results. | ||
| Key Words | ||
| artificial intelligence; concrete columns; deterioration components; machine learning; moment-rotation curves; stacking model | ||
| Address | ||
| A.Khoshkroodi and H.Parvini Sani: Department of Civil Engineering, Zanjan Branch, Islamic Azad University, Zanjan, Iran M.Sadeghi: Centre for Mathematical Plasma Astrophysics (CmPA), KU Leuven, Celestijnenlaan 200B bus 2400, B-3001 Leuven, Belgium | ||