Evaluating the mechanical strength prediction performances of fly ash-based MPC mortar with artificial intelligence approaches

Introduction of Fly ash (FA) in the magnesium phosphate cement (MPC) mortars is considered as sustainable way to advance the microstructural characteristics and reduce the manufacturing cost of MPC products. However, artificial intelligence (AI) approaches are still need to forecast the strength properties of MPC compositions blended with FA and estimate the governing input elements for appropriate mix design with suitable contents. For this aims, the current research elected five AI models based on deep neural network (DNN), optimizable gaussian process regressor (OGPR) and gene expression programming (GEP) to judge the prediction accuracy of mechanical strength values of the MPC-FA compounds, where the literature data was collected for training the models. In addition, laboratory tests were conducted in this study for producing the data and validating the recommended AI methods. As is observed, DNN2 having 3 hidden layer and Bayesian optimization based Gaussian process regressor techniques presented prediction skills above 95% with errors below 5% at the training and validation phases. Moreover, sensitivity analysis of each input variable revealed that FA content has the prime impact on strength achievement of MPC-FA mixtures, which was corroborated by the correlation analysis between inputs and outputs of whole data points. Finally, forecasting the mechanical strength properties of FAbased MPC mortars using the DNN2 and OGPR methods might be applied in the practical field for reducing the workload, labor and material ingesting through optimizing the mix combinations.

Automated summary

This study uses AI models to predict the mechanical strength of MPC-FA compounds, and the results show that the DNN2 and OGPR methods have high prediction accuracy. Sensitivity analysis reveals that the FA content has the main impact on the strength of MPC-FA mixtures. These predictions can be applied in practical fields to reduce workload, labor, and material consumption through optimizing mix combinations.