Neuro-fuzzy modeling of parameters of biocomfort in low-rise buildings

Abstract

The method of creating the mathematical models on the basis of synthetic set of data by the using a neuro-fuzzy modeling is described. The main factors of residential environment, which influence on feeling the comfort and health of man and are evaluated on the basis of index of level of biocomfort, are defined. The set of data is created using experimental data, conclusions of experts and results of mathematical modeling. The features of using neuro-fuzzy T-controller and neural generator of formulas, which provide the possibility of combining synthetic data to develop models are considered. Mathematical models for the calculation of integral indexes of comfort in a dwelling that allowed improve accuracy of evaluation of quality parameters of biocomfort are created. The results of evaluating the index of level of biocomfort by using of created models are presented.