Determine the heat demand of existing buildings with machine learning
Loading...
Authors
Author (Corporation)
Institut Nachhaltigkeit und Energie am Bau, Hochschule für Architektur, Bau und Geomatik FHNW
Publication date
01.12.2023
Typ of student thesis
Course of study
Collections
Type
01A - Journal article
Editors
Editor (Corporation)
Supervisor
Parent work
Journal of Physics: Conference Series
Special issue
CISBAT 2023
DOI of the original publication
Link
Series
Series number
Volume
2600
Issue / Number
3
Pages / Duration
1-7
Patent number
Publisher / Publishing institution
IOP Publishing
Place of publication / Event location
Edition
Version
Programming language
Assignee
Practice partner / Client
Abstract
The renovation rate of existing buildings plays a major role in the Swiss Energy Strategy 2050+. To increase this rate, there must be a simple and cost-effective method to determine the heat demand of existing buildings. In this paper, the generation of such a method, based on the Swiss cantonal building energy certificate (GEAK) database with the help of machine learning (ML), is studied. The aim of the project was to develop a ML model which allows the heat demand of existing buildings to be determined quickly with a minimal set of parameters. The comparison of the GEAK building envelope class for single family houses calculated with the new ML model and the original GEAK classes shows that approximately 62 % have the same class, 32 % differ by one class and 6 % by two classes. The ML model is a good starting point for further refinements and developments.
Keywords
Machine learning, Deep neural network, Head demand
Subject (DDC)
624 - Ingenieurbau und Umwelttechnik
Event
Exhibition start date
Exhibition end date
Conference start date
Conference end date
Date of the last check
ISBN
ISSN
1742-6588
1742-6596
1742-6596
Language
English
Created during FHNW affiliation
Yes
Strategic action fields FHNW
Publication status
Published
Review
Peer review of the complete publication
Open access category
Green
License
Citation
HOFMANN, Joachim Werner, Christian AMOSER, Achim GEISSLER und Monika HALL, 2023. Determine the heat demand of existing buildings with machine learning. Journal of Physics: Conference Series. 1 Dezember 2023. Bd. 2600, Nr. 3, S. 1–7. DOI 10.1088/1742-6596/2600/3/032013. Verfügbar unter: https://doi.org/10.26041/fhnw-5920