Elephant Herding Optimization Based Neural Network to Predict Elastic Modulus of Concrete

摘要

The concrete requirement is proliferating, especially in developing countries. The elastic modulus of concrete plays a vital role as a design parameter in construction applications. The experimental procedures for finding the elastic modulus of concrete are quite complicated and expensive. Thus, researchers have always been looking for better and efficient methods to replace the traditional methods. In the present study, the issue is addressed by hybridized soft computing technique called Elephant Herding Optimization Based Artificial Neural Network (EHO-ANN). The developed model is then validated with linear regression and standard empirical formulas used for estimation of elastic modulus of concrete. The performances are evaluated by statistic measures like Correlation Coefficient (CC), Root Mean Square Error (RMSE) and Mean Absolute Percent Error (MAPE). The developed EHO-ANN model (Train CC 0.9102 and Test CC 0.9095) has outperformed linear regression (Train CC 0.8287 and Test CC 0.7575) and empirical formula (Train CC 0.8170 and Test CC 0.7645) in predicting the elastic modulus of concrete. Hence, the EHO-ANN model can be used as an alternative method to predict the elastic modulus of concrete.