Fuzzy logic-based model for prediction of building wall humidity level

Abstract

This paper presents the viability of using fuzzy logic expert system for the prediction of building wall’s relative humidity. In comparison with traditional logic model, fuzzy logic is more efficient in linking the multiple units to a single output and is invaluable supplements to classical hard computing techniques. The purpose of this work was to investigate the relationship between wall material composition and climate condition, as well as to evaluate the viability of using fuzzy logic expert system in predicting the building wall’s humidity level. The experimental data was taken at different time of the day for different material composition. A fuzzy logic expert system model based on Mamdani approach was subsequently developed to predict the building wall’s humidity level. Verification of the developed fuzzy logic model was carried out through various numerical error criteria. The study revealed that the developed fuzzy system was able to predict the humidity level in the wall material within high accuracy with relative error of less than the acceptable limits of 10% and a goodness of fit very close to unity.