Novel Hybrid RBFNN Machine-Learning Method with Covid-19 Algorithm to Predict Compressive Strength of FRP-Confined Concrete Columns

This paper investigates the effectiveness of two different machine-learning methods for predicting the ultimate strength of rectangular Concrete Columns confined with fiber-reinforced polymer (FRP) sheets. The two machine learning methods are the Radial Basis Function Neural Network (RBFNN) and the RBFNN Hybridized with the Covid-19 Pandemic Optimization algorithm (RBFNN-CPO). The models were compared over the measurements of the Root Mean Square Error (RMSE), Standard Deviation (SD), and correlation coefficient criteria. RBFNN and RBFNN -CPO results were compared with a wide range of experimental data, including 532 samples collected for square and rectangular columns confined by various FRP sheets. Their agreeable globality and consistency demonstrated the ability of RBFNN and RBFNN-CPO to estimate the compressive strength of concrete confined by FRP. In addition, comparing the correlation coefficients for these two models showed that CPO enhanced the performance of RBFNN.

Novel Hybrid RBFNN Machine-Learning Method with Covid-19 Algorithm to Predict Compressive Strength of FRP-Confined Concrete Columns Keywords:

FRP , Concrete Columns , Prediction , Optimization , RBFNN , Covid 19.