| id |
ascaad2023_067 |
| authors |
Yuceses, Birsen; Selçuk, Semra |
| year |
2023 |
| title |
State-of-the-art on Research and Applications of Machine Learning in Building Energy Performance Prediction |
| source |
C+++: Computation, Culture, and Context Proceedings of the 11th International Conference of the Arab Society for Computation in Architecture, Art and Design (ASCAAD), University of Petra, Amman, Jordan [Hybrid Conference] 7-9 November 2023, pp. 275-292. |
| summary |
The construction sector is responsible for 40% of the total energy consumption. The parameters that affect the energy performance of buildings include heating-cooling systems, ventilation systems, lighting, and the implementation of the Energy Performance of Buildings Directive (EPBD) by the European Union. Several regulations and certification systems like PassivHaus, LEED, BREEAM etc. have been developed to regulate this process. There is often a significant discrepancy between the estimated (calculated) energy performance of buildings and the actual energy usage after buildings are operational. This study reviews the efforts made within the context of machine learning to address this 'performance gap'. The systematic literature review focuses on machine learning-based approaches in the context of building energy performance. The areas of focus include reducing the gap between building energy performance predictions and the values obtained during the building's life cycle, and the dynamic updating of the building energy model based on data. The study highlights trends related to the use of machine learning in predicting building energy performance. These trends include the types of buildings considered, the algorithms used, the data utilized, and the targeted areas of energy performance. To conduct the study, a comprehensive literature review was performed using the PRISMA method, which yielded 325 relevant articles out of 3078 articles extracted from the Web of Science database. A bibliometric analysis was then carried out on these 325 articles. From the selected 325 articles, 30 high-impact articles were subjected to content analysis to evaluate potential gaps in the field. |
| series |
ASCAAD |
| email |
|
| full text |
 file.pdf (737,676 bytes) |
| references |
Content-type: text/plain
|
| last changed |
2024/02/13 14:40 |
|
