Synergistic Collaboration of Motion-Based Metaheuristics for the Strength Prediction of Cement-Based Mortar Materials Using TSK Model

Considering mortar material’s extensive use in construction over the last few decades, a robust and reliable method is required to estimate its strength based on mix parameters. The reason behind this is the complex and nonlinear connection between the compressive strength of mortar and its constituent components. This research investigates the utilization of artificial intelligence methods to predict the compressive strength of cement-based mortar materials, both with and without metakaolin. A surrogate model, specifically the Takagi-Sugeno-Kang (TSK) model, was created to predict the compressive strength of mortars based on existing experimental data found in the literature. The findings demonstrate that the TSK model can effectively and reliably estimate the compressive strength of mortars. To prevent the model from overfitting to the existing data during the validation process with new data, it was optimized using the mean square error (MSE) criterion. To this end, a collaborative motion-based algorithm incorporating charged system search (CSS) and colliding bodies optimization (CBO) was employed. Consequently, the developed TSK model is presented as the most suitable predictive technique for addressing the issue of compressive strength prediction in mortars.