Estimation of Heating and Cooling Loads with the Help of Rule-Based Method in Energy Efficient Building Design

The building sector in Turkey is responsible for about %34 of the energy consumption of the country. The energy demand of buildings plays an important role in energy-efficient building design. Machine learning techniques have been used for energy forecasting on buildings. The aim of this study is to develop a rule-based method to explore the effects of building design parameters on heating and cooling loads and to serve energy efficiency. Firsty, the feature selection was applied to the data set from the literature. The heating and cooling load for each parameter were calculated and ranked according to their effects. Then, rule-based models were created by the decision tree. The R performance values of the developed heating and cooling load models are 0.92 and 0.91, respectively. The proposed rule-based method for heating and cooling load is feasible practically as a simple and understandable approach.