Prediction of Stress-strain Relationships for Reinforced Concrete Sections by Implementing Neural Network Techniques

Abstract

The application of neural networks for predicting the stress-strain relationships of reinforced concrete sections is presented. Computation algorithms in the form of numerical analysis were performed on reinforced concrete sections to simulate existing experimental data. A systematic approach is provided by implementing neural networks in the form of prediction by backpropagation algorithms. The efficiency of neural network techniques is demonstrated by means of reconstructing previous experimental work and evaluating several parameters based on neural networks which are in agreement with experimental results. The procedure establishes valid mathematical relationships without relying on a particular algorithm and depends entirely on the manipulation of numerical data.