CNN-based Quick Energy Prediction Model using Image Analysis for Shape Information

Research output: Chapter in book/report/conference proceeding › Conference contribution › Research

Authors

  • Philipp Florian Geyer
  • Manav Mahan Singh

External Research Organisations

  • KU Leuven
  • Technische Universität Berlin
View graph of relations

Details

Original languageEnglish
Title of host publicationProceedings of Building Simulation 2021
Subtitle of host publication17th Conference of IBPSA
EditorsDirk Saelens, Jelle Laverge, Wim Boydens, Lieve Helsen
Pages1311-1316
Number of pages6
ISBN (electronic)978-1-7750520-2-9
Publication statusPublished - Sept 2021
Externally publishedYes
Event 17th IBPSA Conference - Bruges, Belgium
Duration: 1 Sept 2021 → 3 Sept 2021

Publication series

NameBuilding Simulation Conference Proceedings
ISSN (Print)2522-2708

Abstract

Data-driven approaches are useful to substitute computationally expensive tools used for energy prediction. These approaches allow developing quick energy prediction models, essential to promote energy analysis at the early stages. However, developing a wellgeneralising model is challenging due to varying building shape. This article develops such a model using a deep learning approach. A convolutional neural network (CNN) with modified architecture is used to capture the shape and technical specifications. The model predicts energy use intensity with a mean-absolute-percentageerror of 1.51% and root-mean-square-error of 1.06 kWh/m2a. Integrated with building information modelling (BIM), the model predicts probabilistic energy performance for 5000 samples in 65 seconds to support informed decision-making under uncertain scenario.

ASJC Scopus subject areas

Cite this

CNN-based Quick Energy Prediction Model using Image Analysis for Shape Information. / Geyer, Philipp Florian; Singh, Manav Mahan.
Proceedings of Building Simulation 2021: 17th Conference of IBPSA. ed. / Dirk Saelens; Jelle Laverge; Wim Boydens; Lieve Helsen. 2021. p. 1311-1316 (Building Simulation Conference Proceedings).

Research output: Chapter in book/report/conference proceeding › Conference contribution › Research

Geyer, PF & Singh, MM 2021, CNN-based Quick Energy Prediction Model using Image Analysis for Shape Information. in D Saelens, J Laverge, W Boydens & L Helsen (eds), Proceedings of Building Simulation 2021: 17th Conference of IBPSA. Building Simulation Conference Proceedings, pp. 1311-1316, 17th IBPSA Conference, Belgium, 1 Sept 2021. https://doi.org/10.26868/25222708.2021.30250
Geyer, P. F., & Singh, M. M. (2021). CNN-based Quick Energy Prediction Model using Image Analysis for Shape Information. In D. Saelens, J. Laverge, W. Boydens, & L. Helsen (Eds.), Proceedings of Building Simulation 2021: 17th Conference of IBPSA (pp. 1311-1316). (Building Simulation Conference Proceedings). https://doi.org/10.26868/25222708.2021.30250
Geyer PF, Singh MM. CNN-based Quick Energy Prediction Model using Image Analysis for Shape Information. In Saelens D, Laverge J, Boydens W, Helsen L, editors, Proceedings of Building Simulation 2021: 17th Conference of IBPSA. 2021. p. 1311-1316. (Building Simulation Conference Proceedings). doi: 10.26868/25222708.2021.30250
Geyer, Philipp Florian ; Singh, Manav Mahan. / CNN-based Quick Energy Prediction Model using Image Analysis for Shape Information. Proceedings of Building Simulation 2021: 17th Conference of IBPSA. editor / Dirk Saelens ; Jelle Laverge ; Wim Boydens ; Lieve Helsen. 2021. pp. 1311-1316 (Building Simulation Conference Proceedings).
Download
@inproceedings{b48c6d04e6334eaaa4c0876f67892ac1,
title = "CNN-based Quick Energy Prediction Model using Image Analysis for Shape Information",
abstract = "Data-driven approaches are useful to substitute computationally expensive tools used for energy prediction. These approaches allow developing quick energy prediction models, essential to promote energy analysis at the early stages. However, developing a wellgeneralising model is challenging due to varying building shape. This article develops such a model using a deep learning approach. A convolutional neural network (CNN) with modified architecture is used to capture the shape and technical specifications. The model predicts energy use intensity with a mean-absolute-percentageerror of 1.51% and root-mean-square-error of 1.06 kWh/m2a. Integrated with building information modelling (BIM), the model predicts probabilistic energy performance for 5000 samples in 65 seconds to support informed decision-making under uncertain scenario. ",
author = "Geyer, {Philipp Florian} and Singh, {Manav Mahan}",
note = "Publisher Copyright: {\textcopyright} International Building Performance Simulation Association, 2022; 17th IBPSA Conference ; Conference date: 01-09-2021 Through 03-09-2021",
year = "2021",
month = sep,
doi = "10.26868/25222708.2021.30250",
language = "English",
series = "Building Simulation Conference Proceedings",
pages = "1311--1316",
editor = "Dirk Saelens and Jelle Laverge and Wim Boydens and Lieve Helsen",
booktitle = "Proceedings of Building Simulation 2021",

}

Download

TY - GEN

T1 - CNN-based Quick Energy Prediction Model using Image Analysis for Shape Information

AU - Geyer, Philipp Florian

AU - Singh, Manav Mahan

N1 - Publisher Copyright: © International Building Performance Simulation Association, 2022

PY - 2021/9

Y1 - 2021/9

N2 - Data-driven approaches are useful to substitute computationally expensive tools used for energy prediction. These approaches allow developing quick energy prediction models, essential to promote energy analysis at the early stages. However, developing a wellgeneralising model is challenging due to varying building shape. This article develops such a model using a deep learning approach. A convolutional neural network (CNN) with modified architecture is used to capture the shape and technical specifications. The model predicts energy use intensity with a mean-absolute-percentageerror of 1.51% and root-mean-square-error of 1.06 kWh/m2a. Integrated with building information modelling (BIM), the model predicts probabilistic energy performance for 5000 samples in 65 seconds to support informed decision-making under uncertain scenario.

AB - Data-driven approaches are useful to substitute computationally expensive tools used for energy prediction. These approaches allow developing quick energy prediction models, essential to promote energy analysis at the early stages. However, developing a wellgeneralising model is challenging due to varying building shape. This article develops such a model using a deep learning approach. A convolutional neural network (CNN) with modified architecture is used to capture the shape and technical specifications. The model predicts energy use intensity with a mean-absolute-percentageerror of 1.51% and root-mean-square-error of 1.06 kWh/m2a. Integrated with building information modelling (BIM), the model predicts probabilistic energy performance for 5000 samples in 65 seconds to support informed decision-making under uncertain scenario.

UR - http://www.scopus.com/inward/record.url?scp=85151536066&partnerID=8YFLogxK

U2 - 10.26868/25222708.2021.30250

DO - 10.26868/25222708.2021.30250

M3 - Conference contribution

T3 - Building Simulation Conference Proceedings

SP - 1311

EP - 1316

BT - Proceedings of Building Simulation 2021

A2 - Saelens, Dirk

A2 - Laverge, Jelle

A2 - Boydens, Wim

A2 - Helsen, Lieve

T2 - 17th IBPSA Conference

Y2 - 1 September 2021 through 3 September 2021

ER -