Abstract
The application of machine learning (ML) modelling in daylight prediction has been a promising approach for reliable and effective visual comfort assessment. Although many advancements have been made, no standardized ML modelling framework exists in daylight assessment. In this study, 625 different building layouts were generated to model useful daylight illuminance (UDI). Two state-of-the-art ML algorithms, eXtreme Gradient Boosting (XGBoost) and random forest (RF), were employed to analyze UDI in four categories: UDI-f (fell short), UDI-s (supplementary), UDI-a (autonomous), and UDI-e (exceeded). A feature (internal finish) was introduced to the framework to better reflect real-world representation. The results show that XGBoost models predict UDI with a maximum accuracy of R2 = 0.992. Compared to RF, the XGBoost ML models can significantly reduce prediction errors. Future research directions have been specified to advance the proposed framework by introducing new features and exploring new ML architectures to standardize ML applications in daylight prediction.
Article PDF
Similar content being viewed by others
Data availability
Data will be made available on reasonable request.
Abbreviations
- ANN:
-
artificial neural network
- d :
-
distance from a light sensor to a corner of a window
- I :
-
internal finish
- MAE:
-
mean absolute error
- ML:
-
machine learning
- MLM:
-
machine learning model
- RF:
-
random forest
- RMSE:
-
root mean square error
- R 2 :
-
coefficient of determination
- UDI:
-
useful daylight illuminance
- UDI-a :
-
useful daylight illuminance (autonomous)
- UDI-e :
-
useful daylight illuminance (exceeded)
- UDI-f :
-
useful daylight illuminance (fell-short)
- UDI-s :
-
useful daylight illuminance (supplementary)
- w :
-
rotation of a window with correspondence to a light sensor
- x :
-
distance from a light sensor to a perimeter obstacle
- XGBoost:
-
eXtreme Gradient Boosting
References
Alsharif R, Arashpour M, Golafshani EM, et al. (2022). Machine learning-based analysis of occupant-centric aspects: critical elements in the energy consumption of residential buildings. Journal of Building Engineering, 46: 103846.
Arashpour M, Ngo T, Li H (2021). Scene understanding in construction and buildings using image processing methods: a comprehensive review and a case study. Journal of Building Engineering, 33: 101672.
Arashpour M, Kamat V, Heidarpour A, et al. (2022). Computer vision for anatomical analysis of equipment in civil infrastructure projects: Theorizing the development of regression-based deep neural networks. Automation in Construction, 137: 104193.
Arashpour M (2023). AI explainability framework for environmental management research. Journal of Environmental Management, 342: 118149.
Arashpour M, Golafshani EM, Parthiban R, et al. (2023). Predicting individual learning performance using machine-learning hybridized with the teaching-learning-based optimization. Computer Applications in Engineering Education, 31: 83–99.
Arbab M, Rahbar M, Arbab M (2021). A comparative study of artificial intelligence models for predicting interior illuminance. Applied Artificial Intelligence, 35: 373–392.
Ayoub M (2020). A review on machine learning algorithms to predict daylighting inside buildings. Solar Energy, 202: 249–275.
Belitz K, Stackelberg PE (2021). Evaluation of six methods for correcting bias in estimates from ensemble tree machine learning regression models. Environmental Modelling & Software, 139: 105006.
Blackwell B (2002). Light exposure to sensitive artworks during digital photography. Spectra, 26(2): 24–28.
Brembilla E, Drosou NC, Mardaljevic J (2022). Assessing daylight performance in use: A comparison between long-term daylight measurements and simulations. Energy and Buildings, 262: 111989.
Carlucci S, Causone F, De Rosa F, et al. (2015). A review of indices for assessing visual comfort with a view to their use in optimization processes to support building integrated design. Renewable and Sustainable Energy Reviews, 47: 1016–1033.
Chen T, Guestrin C (2016). XGBoost: A scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA.
Chen Y, Chen B, Deng J, et al. (2022). The integration model of objective and subjective data of residential indoor environment quality in Northeast China based on structural equation modeling. Building Simulation, 15: 741–754.
Chi DA (2022). Solar energy density as a benchmark to improve daylight availability and energy performance in buildings: A single metric for a single-objective optimization. Solar Energy, 234: 304–318.
Davoodi A, Johansson P, Aries M (2020). The use of lighting simulation in the evidence-based design process: A case study approach using visual comfort analysis in offices. Building Simulation, 13: 141–153.
Day JK, Futrell B, Cox R, et al. (2019). Blinded by the light: Occupant perceptions and visual comfort assessments of three dynamic daylight control systems and shading strategies. Building and Environment, 154: 107–121.
Fang J, Zhao Y, Tian Z, et al. (2022). Analysis of dynamic louver control with prism redirecting fenestrations for office daylighting optimization. Energy and Buildings, 262: 112019.
Ghobad L, Glumac S (2018). Daylighting and energy simulation workflow in performance-based building simulation tools. In: Proceedings of the 2018 Building Performance Analysis Conference and Simbuild, Chicago, IL, USA.
González S, García S, Del Ser J, et al. (2020). A practical tutorial on bagging and boosting based ensembles for machine learning: Algorithms, software tools, performance study, practical perspectives and opportunities. Information Fusion, 64: 205–237.
Gunay HB, O’Brien W, Beausoleil-Morrison I, et al. (2017). Development and implementation of an adaptive lighting and blinds control algorithm. Building and Environment, 113: 185–199.
He Q, Li Z, Gao W, et al. (2021). Predictive models for daylight performance of general floorplans based on CNN and GAN: A proof-of-concept study. Building and Environment, 206: 108346.
Khidmat RP, Fukuda H, Paramita B, et al. (2022). Investigation into the daylight performance of expanded-metal shading through parametric design and multi-objective optimisation in Japan. Journal of Building Engineering, 51: 104241.
Le-Thanh L, Nguyen-Thi-Viet H, Lee J, et al. (2022). Machine learning-based real-time daylight analysis in buildings. Journal of Building Engineering, 52: 104374.
Lin C-H, Tsay Y-S (2021). A metamodel based on intermediary features for daylight performance prediction of façade design. Building and Environment, 206: 108371.
Liu G, Qu G, Ren L, et al. (2022). The influence mechanism of daylight visual evaluation in college classrooms under visual field physiological characteristics of student group: Case study. Building and Environment, 209: 108655.
Lou S, Li DHW, Lam JC, et al. (2016). Prediction of diffuse solar irradiance using machine learning and multivariable regression. Applied Energy, 181: 367–374.
Luo Z, Sun C, Dong Q, et al. (2022). Key control variables affecting interior visual comfort for automated louver control in open-plan office—A study using machine learning. Building and Environment, 207: 108565.
Lv Y, Peng H, He M, et al. (2019). Definition of typical commercial building for South China’s Pearl River Delta: local data statistics and model development. Energy and Buildings, 190: 119–131.
Manfren M, James PA, Tronchin L (2022). Data-driven building energy modelling - An analysis of the potential for generalisation through interpretable machine learning. Renewable and Sustainable Energy Reviews, 167: 112686.
Mardaljevic J, Andersen M, Roy N, et al. (2012). Daylighting metrics: Is there a relation between useful daylight illuminance and daylight glare probabilty? In: Proceedings of the 1st Building Simulation and Optimization Conference, Loughborough, UK.
Michael A, Heracleous C (2017). Assessment of natural lighting performance and visual comfort of educational architecture in Southern Europe: The case of typical educational school premises in Cyprus. Energy and Buildings, 140: 443–457.
Montaser Koohsari A, Heidari S (2022). Subdivided venetian blind control strategies considering visual satisfaction of occupants, daylight metrics, and energy analyses. Energy and Buildings, 257: 111767.
Nabil A, Mardaljevic J (2005). Useful daylight illuminance: a new paradigm for assessing daylight in buildings. Lighting Research & Technology, 37: 41–57.
Ngarambe J, Adilkhanova I, Uwiragiye B, et al. (2022). A review on the current usage of machine learning tools for daylighting design and control. Building and Environment, 223: 109507
Oyedele A, Ajayi A, Oyedele LO, et al. (2021). Deep learning and Boosted trees for injuries prediction in power infrastructure projects. Applied Soft Computing, 110: 107587.
Papadopoulos S, Azar E, Woon WL, et al. (2018). Evaluation of tree-based ensemble learning algorithms for building energy performance estimation. Journal of Building Performance Simulation, 11: 322–332.
Park Y, Ho JC (2021). Tackling overfitting in boosting for noisy healthcare data. IEEE Transactions on Knowledge and Data Engineering, 33: 2995–3006.
Peng H, Li M, Lou S, et al. (2020). Investigation on spatial distribution and thermal properties of typical residential buildings in South China’s Pearl River Delta. Energy and Buildings, 206: 109555.
Pérez-Fargallo A, Rubio-Manzano C, Martínez-Rocamora A, et al. (2018). Linguistic descriptions of thermal comfort data for buildings: Definition, implementation and evaluation. Building Simulation, 11: 1095–1108.
Shekar BH, Dagnew G (2019). Grid search-based hyperparameter tuning and classification of microarray cancer data. In: Proceedings of 2019 Second International Conference on Advanced Computational and Communication Paradigms (ICACCP), Gangtok, India.
Shi L, Zhang Y, Wang Z, et al. (2021). Luminance parameter thresholds for user visual comfort under daylight conditions from subjective responses and physiological measurements in a gymnasium. Building and Environment, 205: 108187.
Thrampoulidis E, Mavromatidis G, Lucchi A, Orehounig K (2021). A machine learning-based surrogate model to approximate optimal building retrofit solutions. Applied Energy, 281: 116024.
Veloso B, Gama J, Malheiro B, et al. (2021). Hyperparameter self-tuning for data streams. Information Fusion, 76: 75–86.
Wagiman KR, Abdullah MN, Hassan MY, et al. (2021). A new metric for optimal visual comfort and energy efficiency of building lighting system considering daylight using multi-objective particle swarm optimization. Journal of Building Engineering, 43: 102525.
Wang Z, Yu H, Luo M, et al. (2019). Predicting older People’s thermal sensation in building environment through a machine learning approach: Modelling, interpretation, and application. Building and Environment, 161: 106231.
Yang L, Shami A (2020). On hyperparameter optimization of machine learning algorithms: theory and practice. Neurocomputing, 415: 295–316.
Yngvesson L, Adolfsson E (2018). The impact of scale when using models of daylight analysis. Jönköping University, Sweden.
Zhang K, Yang J, Sha J, et al. (2022). Dynamic slow feature analysis and random forest for subway indoor air quality modeling. Building and Environment, 213: 108876.
Acknowledgements
The authors are grateful for support from the Australian Research Council (ARC) through the Linkage Infrastructure, Equipment and Facilities (LE210100019). The assistance of the ASCII Lab members at Monash University is greatly appreciated.
Funding
Open Access funding enabled and organized by CAUL and its Member Institutions.
Author information
Authors and Affiliations
Contributions
Alsharif R., Arashpour M., and Golafshani E. conceived and planned the experiments. Alsharif R., Arashpour M., and Golafshani E. carried out the experiments. Alsharif R. and Golafshani E. planned and carried out the machine learning phase. Alsharif R. and Arashpour M. contributed to sample preparation. Alsharif R., Arashpour M., Golafshani E., Bazli M., and Mohandes S. contributed to the interpretation of the results. Alsharif R. took the lead in writing the manuscript. All authors provided critical feedback and helped shape the research, analysis, and manuscript.
Corresponding author
Ethics declarations
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Rights and permissions
This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/
About this article
Cite this article
Alsharif, R., Arashpour, M., Golafshani, E. et al. Ensemble machine learning framework for daylight modelling of various building layouts. Build. Simul. 16, 2049–2061 (2023). https://doi.org/10.1007/s12273-023-1045-x
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12273-023-1045-x