Journal
STRUCTURES
Volume 30, Issue -, Pages 692-709Publisher
ELSEVIER SCIENCE INC
DOI: 10.1016/j.istruc.2021.01.036
Keywords
Axial compression capacity (ACC); Machine learning; Gene expression programing (GEP); Square concrete-filled steel tubular (SCFST) columns; Empirical correlations; External validation
Categories
Ask authors/readers for more resources
The study proposed a new framework using Gene Expression Programing (GEP) to develop novel models with closed-form equations to describe the behavior of axial compression capacity (ACC) for Square Concrete-Filled Steel Tubular (SCFST) columns. Six novel predictive formulas were proposed based on the GEP-approach for an accurate ACC modeling. The results showed that the GEP-based formulations outperformed existing codes and correlations, demonstrating the efficiency of GEP in deriving new models for modeling the complex behavior of ACC for SCFST columns.
Axial compression capacity (ACC) is an important parameter for the concrete-filled steel tubular columns to measure the resistance of these fundamental elements, which used in the construction of several structures types. For this purpose, the Gene Expression Programing (GEP) is proposed in this paper as a new framework for the development of novel models with closed-form equations to describe the behavior of the axial compression capacity (ACC) for Square Concrete-Filled Steel Tubular (SCFST) columns. For an accurate ACC modeling, six novel predictive formulas based on the GEP-approach were proposed by incorporating different combinations of the input variables. These latter were obtained from a large dataset that includes 300 experimental tests with different ranges and varieties. Besides, the most known codes and empirical correlations for modeling the behavior of ACC for SCFST columns were reviewed, whereas the performance, accuracy, and efficiency of the proposed models and the excited codes and correlations were investigated and compared using several statistical criteria and graphical illustration. Results show that the best explicit closed-form correlation extracted based on the GEP-approach exhibit an overall coefficient of determination (R-2) value of 0.9943. Furthermore, the outcome results indicate that the efficiency of the proposed GEP-based formulations outperform the excited codes and correlations, which proves that the GEP is a powerful technique to derive a new model for modeling the complex behavior of the ACC for SCFST columns.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available