Perspectiva de Uso da Inteligência Artificial (IA) para a Eficiência Energética em Prédios Públicos

Justino, Marcelo Pereira; Silva, Fernando Selleri; Rabêlo, Olivan da Silva

doi:10.9771/cp.v13i3.33079

Cited by 2 publications

(1 citation statement)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…According to Justino, Silva and Silva Rabelo (2020) In the international sphere, there is a consolidated scenery related to the studies development with the use of machine learning in predictions of energy consumption in buildings. In Tsanas and Xifara (2012), a set of simulated data by Ecotec software referred to the parameters that had a profound influence of an edification in the acquisition of the heating and cooling, are submitted to two learning machine methods, known as, linear regression and random forest .…”

Section: Introductionmentioning

confidence: 99%

Aplicação de um modelo de aprendizado de máquina em estudo de eficiência energética de edificações: foco para sistemas construtivos leves

Moro

Martins²,

Giglio³

2022

Semina: Ciênc. Ex. Tech.

View full text Add to dashboard Cite

The use of machine learning techniques in thermoenergetic performance studies of buildings emerges as an alternative to conventional methods which analysis require greater data complexity. This research aims to apply a machine learning technique in the study of energy efficiency of a building executed in a light construction system. Thus, an algorithm was implemented referring to an optimized model of classification and regression tree (CART) for application in a data set. This data set includes 2048 parametric simulations of a housing in a light construction system to the climate of the city of São Paulo, whose output indicators are the annual thermal load for heating and the annual thermal load for cooling. From the application of a tree pruning methodology and the use of Grid Search and k-fold Cross Validation techniques, the training and testing of the model was repeated 100 times, thus obtaining average results of 1.11\% of error for heating loads and 1.52\% of error for predicting cooling loads. Subsequently, a sensitivity analysis was performed, revealing the thermal transmittance property of the walls as the parameter with the greatest influence on the prediction of heating load and the condition of contact between the ground and the floor as the parameter with the greatest influence on the prediction of cooling load. Finally, decision trees were generated for visual analysis of strategies that can be adopted to obtain better levels of thermoenergetic performance. Thus, a more simplified diagnosis of energy efficiency was obtained, with low complexity in the interpretation of its results, favoring greater diffusion of the technology in light systems.

show abstract

Section: Introductionmentioning

confidence: 99%