Artificial intelligence based prediction model for the long-term heat flux losses through ground applied to large non-residential buildings

One of the most important directions towards global CO2 emissions and primary energy consumption reduction is to increase energy efficiency in the sector of residential and non-residential buildings. The evaluation of building envelope heat losses through ground, as part of the building energy demand and energy consumption, it still has a lack of comprehensive knowledge relative to the large buildings. This article aims to use artificial neural networks (ANNs) to allow long-term prediction of the heat transfer losses through ground during heating season for the large dimensions slabs, which are specific for many non-residential buildings, in order to reduce the significant resources needed for the numerical analysis in time-dependent state. A hybrid approach is proposed by developing an application to study a less investigated area of civil engineering.