Abstract: In this paper the authors discuss some useful procedures for energy efficiency in buildings. For this purpose, it is necessary to simulate the temperature trend concerning a building. More specifically, a hybrid solar/photovoltaic system suitable for smart grid is considered, in order to define some generic critical aspects concerning energy efficiency. Starting from critical points, such as the analysis of temperature ripples and heat leakage, the authors provide a workflow procedure in order to plan the monitoring activities aimed to reduce energy losses. So, a part of this procedure involves the thermal imaging analysis. The authors focused the attention on a single thermal image, divided in 15 subimages through the sliding windows approach, to predict the energy leakage around a window. For the experimental and numerical evaluation have been used both the k-Means algorithm for clustering and the Nearest Neighbor (NN) classifier to represent a particular risk map. RapidMiner tool has been used for the processing of the radiometric measurements. The proposed procedure provides important information about heat leakage associated to energy efficiency in building.
Keywords: Building Energy Efficiency, Thermal Imaging Technique, k-Means