APPLICATION OF ARTIFICIAL INTELLIGENCE FOR PREDICTING THE COMPRESSIVE STRENGTH OF CONCRETE USING NATURAL AGGREGATE

Abstract

This seminar presentation explored the application of various artificial intelligence techniques such as Artificial Neural network (ANN), Adaptive Neuro-Fuzzy Inference System (ANFIS) and Multiple Linear Regression (MLR) for predicting the compressive strength of concrete using natural aggregates. Twenty-seven different experimental data points which was augmented to 180 data points was used in the study. The ANN, ANFIS and MLR models were developed, trained, tested and validated with the augmented data using MATLAB software. Statistical evaluators like the R2, MSE and the RMSE was used to evaluate the algorithm with the strongest predictive capability. The results obtained from the analysis revealed distinct performance variations among the three AI models studied. Both the ANN and ANFIS models consistently demonstrated superior predictive capabilities compared to the MLR model. The ANN gave R2 of 1, MSE of 8.66e-26 and RMSE 2.94e-13, the ANFIS gave R2 values of 1, MSE of 0.00033 and RMSE of 0.0183 while the MLR reported R2 values of 0.1243, MSE of 85.93 and RMSE of 9.27. The ANN model was adjudged to be the best prediction model for concrete containing natural aggregate based on the performance metrics.