Artificial Intelligence Eased Method for the Long-Term Heat Losses Calculation Applied to Large Slabs On Ground

Abstract

As part of the efforts to reduce energy consumption and greenhouse gas emissions on a large scale, one of the main research and development directions is focused on the residential and non-residential buildings sector. This paper assesses heat losses of buildings through ground, as part of the energy demand and energy consumption of the building, a domain that still lacks comprehensive knowledge, especially for the large buildings. In order to reduce the significant resources required for numerical analysis in time-dependent state, the authors propose the use of artificial intelligence to allow long-term prediction of the hourly heat transfer losses through ground for large dimension slabs. The attention was directed to this subject due that this slabs are specific for many types of non-residential buildings, such as commercial buildings (hypermarkets, malls), production halls or even educational buildings. An alternative approach is undertaken in the direction of carrying out detailed analyzes using numerical methods to establish input data for neural networks, in order to predict the building's part thermal response, by totally substituting numerical analysis, for arbitrary sizes of slabs on ground.