Prediction of Space Heating Consumption in District Heated Apartments
Enhancing energy efficiency from the producer to the end-users is a prioritized objective of European energy policies. The most difficult element to understand and to control is the last link, the consumer. Most simulation tools for the estimation of heat demand in buildings are based on the input parameters according to environmental conditions and the features of the materials and equipments. Studies have revealed that even if sophisticated programs are used, a gap between calculated and post occupancy space heating consumption rates occurs. Based on billing history, the present paper proposes a cluster analysis method to identify the grouping trends of habitants’ consumption. For modeling the heat consumption at end-user level and predicting the trends of energy demand, artificial neural networks (ANNs) technique was used. Unlike most prediction methods that have the entire building in view, the proposed technique can be applied to individual apartments situated in condominiums.