Fuzzy Model for the Prediction of the Dosage of Cement Mortars

Abstract

In this work the behaviour of cement mortars is studied in the laboratory through experimentation with different dosages. A fuzzy system is designed with the obtained experimental data: linguistic variables and labels, membership functions’ position and rules adjustment are included in a Mamdani inference system. The main goal of the fuzzy system is to model the mortars’ behavior for helping with their manufacturing. Due to the lack of regulations about the mortars’ dosage, the main goal of this work is to provide construction professionals with a tool for helping the mortars’ preparation. The fuzzy logic prediction system is validated by comparing its predictive output with the real values obtained in new experiments carried out in the laboratory.