Modeling Confinement Efficiency of FRP-Confined Concrete Column Using Radial Basis Function Neural Network

The establishment of confined concrete strength is an important issue in fiber reinforced polymer (FRP)-confined concrete column. This paper explores the use of Radial Basis Function Neural Network (RBFNN) in predicting the confinedment efficiency of FRP-confined concrete. Based on 362 experimental datas, the RBFNN model with highly non-linear reflection relationship was found and tested by the experimental data. A comparison study between the RBFNN model and four well-known models is carried out, it was found that the RBFNN model could reasonably capture the underlying behavior of FRP-confined concrete and provide better results than other models. The sensitivity analysis of the influential factor is also discussed, it shows that RBFNN-based modeling is a practical method for predicting the confinement efficiency of FRP-confined concrete.