Artificial Neural Networks Model for Predicting the Strength of FRP-Contained Concrete

Abstract

Numerous studies have demonstrated that FRP (Fibre-reinforced Polymer) can significantly increase the strength of concrete columns. Numerous mathematical equations and manual methods are available for predicting the strength of concrete columns composed of FRP, all of which are time-consuming tasks. This present study develops a novel computerized method for determining the axial strain and axial strength of FRP (Fibre-reinforced Polymer)-confined concrete columns utilizing real-time experimental data and artificial neural networks (ANNs). In order to increase prediction accuracy, an ANN model is trained and evaluated using experimental data collected in real-time. Additionally, advanced pre-processing techniques are applied in this study to minimize noise and enhance the prediction accuracy of the suggested ANN model. To demonstrate the efficacy of this proposed strategy, this model is trained and verified using the data set. The experimental outcomes from training and validation have been compared to recent methods. It is evident from the comparison results that the proposed method has reduced MAE, RSME and regression values.