An Integrated Optimization Model Using Machine Learning for Daylighting, Thermal Comfort, and Energy Efficiency in Historic Townhouses
Light & Engineering · 25.04.2026 · 28
DOI: 10.33383/2025-083 ↗Kurz gefasst
Fachartikel in der Zeitschrift Light & Engineering, online erschienen am 19.04.2026. Verfasst von Nurefsan Sonmez und 3 weiteren Autor:innen. Die Arbeit nennt 9 Literaturangaben.
Abstract (Originalsprache)
Historic buildings pose challenges for daylighting improvements due to preservation constraints, however, improving their performance is essential for climate goals. This study seeks to improve daylighting performance, thermal comfort, and energy efficiency in historic buildings while maintaining their historic value through various design strategies, including glass upgrades, shading elements, and light tube integration, using a multi-objective optimization model. To analyse the performance of the model, two case studies from distinct climates, Mediterranean and Temperate Oceanic, were evaluated through Grasshopper plug-ins, Ladybug and Honeybee, and optimized via Octopus plug-in. A machine learning tool, Dodo plug-in was integrated to speed up the evaluation process and estimate simulation results while exploring a larger number of design alternatives. The optimization results reveal that daylighting strategies improved UDI up to 83.6 % and reduced ASE to 0 %, and improved TCP up to 6.73%, while energy use decreased by up to 8.81 % with shading and light tubes proving most effective in warm climates and roof windows in colder ones. The study confirms that performance-based design can enhance environmental efficiency without compromising buildings' historic values.
Zitieren
APA
Sonmez, N., Kunduracı, A. C., Ugurlu, C., & Erkan, Y. (2026). An Integrated Optimization Model Using Machine Learning for Daylighting, Thermal Comfort, and Energy Efficiency in Historic Townhouses. Light & Engineering, 28. https://doi.org/10.33383/2025-083
Vancouver
Sonmez N, Kunduracı AC, Ugurlu C, Erkan Y. An Integrated Optimization Model Using Machine Learning for Daylighting, Thermal Comfort, and Energy Efficiency in Historic Townhouses. Light & Engineering. 2026:28. doi:10.33383/2025-083
BibTeX
@article{Sonmez2026,
author = {Sonmez, Nurefsan and Kunduracı, Arzu Cılasun and Ugurlu, Cemre and Erkan, Yonca},
title = {An Integrated Optimization Model Using Machine Learning for Daylighting, Thermal Comfort, and Energy Efficiency in Historic Townhouses},
journal = {Light & Engineering},
year = {2026},
pages = {28},
doi = {10.33383/2025-083}
}
EUVIMED