BIM-oriented data mining for thermal performance of prefabricated buildings

Lucianne Casasanta Garcia (1) , Bernard Kamsu-Foguem (2)

1 Universidad Nacional de Rosario [Santa Fe]
2 LGP - Laboratoire Génie de Production

The use of energy efficiency procedures is a typical practice in building construction process that creates a huge amount of data regarding the building. This is particularly valid in structures which include complex collaborations, for example, ventilation, sunlight-based increases, inner additions, and warm mass. This paper proposes a new approach for automating building construction when improving their energy efficiency, aiming to foresee comfort levels based on Heating, Ventilating, Air Conditioning (HVAC), constructive systems performance, environmental conditions, and occupant behavior. More specifically, it presents a research work about thermal performance of prefabricated construction systems developed by an Argentine enterprise called Astori, using two Knowledge Discovery in Databases (KDD) processes to extract knowledge. In this context, Building Information Modeling (BIM) will give data to support the calculation to outline goal levels of a sustainable building performance concerning classification systems. The data were collected from a project in Uruguay referring to the construction systems and the energy efficiency of the building. The data mining tool SPMF was used to test the performance of classification and its use in prediction. Particularly, FP-Growth Algorithm and Clustering methodologies were used to analyze a combination of ambient conditions, in order to compare them using Revit© software. The results generated by these methods can be generalized for a set of buildings, according to the objective to be achieved concerning the thermal building performance.

Mots clés: Data mining; Association rules; Clustering; Building information; Green buildings

Domaines: Autre [cs.OH]; Environnement et Société

Lucianne Casasanta Garcia, Bernard Kamsu-Foguem. BIM-oriented data mining for thermal performance of prefabricated buildings. Ecological Informatics, 2019, 51, pp.61-72. ⟨10.1016/j.ecoinf.2019.02.012⟩. ⟨hal-02106932⟩