Open AccessDeep neural networks as surrogate models for urban energy simulations
Author(s) -
José R. Vázquez-Canteli,
Aysegul Dilsiz Demir,
Julien D. Brown,
Zoltán Nagy
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1343/1/012002
Subject(s) - artificial neural network , efficient energy use , work (physics) , computer science , energy (signal processing) , building energy simulation , deep neural networks , energy demand , climate change , surrogate model , environmental science , simulation , artificial intelligence , energy performance , machine learning , environmental economics , engineering , statistics , geology , mathematics , economics , mechanical engineering , oceanography , electrical engineering
Building energy simulation helps governments implement effective policies to increase energy efficiency. In this work, we use deep neural networks (DNN) to create a surrogate model of an urban energy simulator. We modelled 7,860 buildings, with 2,620 geometries, and simulated them across all the climatic regions of the US. With these 68 million hourly data points, we trained two DNNs to predict the solar gains and thermal losses. The DNNs reduce computational time by a factor of 2500 while maintaining good accuracy (R 2 =0.85). Possible applications are prediction of energy demand due to climate change and building refurbishment measures.