Modeling of thermal comfort in air conditioned rooms by fuzzy regression analysis

Thermal comfort is a vague and subjective term. Thermally comfortable is not only influenced by the physical environment but also influenced by individual feelings and perception, which is completely subjective and is generally expressed in linguistic terms. Traditional statistic approaches cannot handle these subjective aspects effectively. Fuzzy sets appear to be ideally suited for the modeling of this partially subjective system. To illustrate the effectiveness of fuzzy regression, two particularly fuzzy regression approaches were used to model thermal comfort. To obtain the needed data, experiments were first carried out. The influencing factors considered in the experiments included both the environment influences and the individual differences such as metabolic rate. The results are analyzed and the influence of individual feeling or perception plays an important role in the experimental results and in the modeling.