Balancing Composite Motion Optimization and Artificial Neural Network for the Prediction of Critical Load of Concrete-Filled Steel Tubes Under Axial Compression

In this study, the critical capacity of concrete-filled steel tubes (CFST) under axial compression is investigated through a machine learning model (BCM-NN) that is a combination of Balancing Composite Motion Optimization (BCM) and Artificial Neural Network (NN). To develop the model, a database of experimental results is gathered from literature works. Such database includes variables related to the CFST geometry parameters and its mechanical behavior. Quality assessments namely coefficient of determination (R²) and Root Mean Square Error (RMSE) are used to evaluate the performance of the proposed model. The value range of each model parameter has also been studied in this work. The obtained results show that the hybrid BCM-NN model produced great performance in predicting the critical load of CFST compared to conventional neural network algorithms.