Predicting Compression Strength of Reinforced Concrete Columns Confined by FRP Using Meta-Heuristic Methods

There are several methods to predict the compression strength of reinforced concrete columns confined by FRP, such as experimental methods, theory of elasticity and plasticity. Meanwhile, due to its good potential and high accuracy in predicting different problems, the soft computing techniques has attracted considerable attentions. Soft computing includes methods and programs to deal with complex computational problems. The objective of this study is to evaluate and compare the performance of four methods of Least Squares Support Vector Machine (LS-SVM), the Weight Least Squares Support Vector Machine (WLS-SVM), Adaptive Neuro-Fuzzy Inference System (ANFIS) and Particle Swarm Optimization - Adaptive Network based Fuzzy Inference System (PSO-ANFIS) for predicting the compression strength of reinforced concrete columns confined by FRP. A total of 95 laboratory data are selected for use in these methods. The Root Mean Square Error (RMSE) and the correlation coefficient of the results are used to validate and compare the performance of the methods. The results of the study show that the PSO-ANFIS method with the RMSE of 4.610 and the coefficient of determination of R2 = 0.9677 predicts compression strength of reinforced concrete columns confined by FRP with high accuracy and therefore, it can be a good alternative to time-consuming and costly laboratory methods.