Optimization of Horizontal and Vertical Shading Devices for Daylight Performance in Residential Buildings in Isfahan, A Simulation and Machine Learning Approach

Subject Areas : Ecology, climate and energy

Ahmad Najafi ^{1 *} , Mahdi Torkan ² , Mohammad saeid Sobhani ³

1 - Department of Architecture and Urban Planning, National University of Skills (NUS), Tehran, Iran.
2 - Department of Architecture and Urban Planning, National University of Skills (NUS), Tehran, Iran
3 - Department of Architecture and Urban Planning, National University of Skills (NUS), Tehran, Iran

Received: 2025-01-24 Accepted : 2025-03-13 Published : 2025-03-19

Keywords: Optimization of Shading Devices, Daylighting, Machine Learning Algorithms, Lighting Simulation, Visual Comfort,

Abstract :

This study investigates the optimization of horizontal and vertical shading devices to improve daylighting performance in residential buildings in Isfahan, using a combination of simulation methods and machine learning. The primary objective is to balance enhancing visual comfort by maximizing natural light penetration and reducing solar heat gain. Despite the importance of shading devices in reducing cooling loads, previous research has often overlooked their impact on daylight distribution and visual comfort in hot and dry climates. This study addresses this gap by proposing a comprehensive framework. The research methodology consists of three main parts: (1) setting up a reference living room in Isfahan with precise geometric and material specifications, (2) conducting daylighting simulations using Ecotect software (version 2011) to evaluate the performance of shading devices under different seasonal conditions, and (3) employing machine learning models, including multivariate regression, decision trees, random forests, and artificial neural networks, to analyze and optimize the design of shading devices. Genetic algorithms were used to identify optimal configurations. Key findings indicate that the distance between shading elements has the most significant impact on daylight penetration, while the depth and number of shading elements negatively affect daylighting performance. The random forest model, with high predictive accuracy (R² = 0.9679 for vertical shading devices and 0.9877 for horizontal ones), was identified as the best-performing model. Sensitivity analysis also confirmed the importance of distance in optimizing shading devices. The study concludes that optimizing shading devices can significantly improve daylighting performance in residential buildings. It is recommended that designers maximize the distance between shading elements, reduce their depth, and carefully adjust their number. This research demonstrates that integrating machine learning with simulation tools provides a powerful framework for designing sustainable and energy-efficient buildings.

References:

Yi, H., Kim, M.-J., Kim, Y., Kim, S.-S., & Lee, K.-I. (2019). Rapid Simulation of Optimally Responsive Façade during Schematic Design Phases: Use of a New Hybrid Metaheuristic Algorithm. Sustainability, 11(9), 2681. https://doi.org/10.3390/su11092681
Xie, Y., Ji, W., Li, T., Liang, Y., & Li, F. (2019). Location Aided and Machine Learning-Based Beam Allocation for 3D Massive MIMO Systems. 2019 15th International Wireless Communications & Mobile Computing Conference (IWCMC), 836-841. https://doi.org/10.1109/IWCMC.2019.8766724.
Wagle, S., & Paranjape, A. (2020). Use of Simulation-Aided Reinforcement Learning for Optimal Scheduling of Operations in Industrial Plants. 2020 Winter Simulation Conference (WSC), 572-583. https://doi.org/10.1109/WSC48552.2020.9383893.
Nourkojouri, H., Shafavi, N., Tahsildoost, M., & Zomorodian, Z. (2021). Development of a Machine-Learning Framework for Overall Daylight and Visual Comfort Assessment in Early Design Stages. Journal of Daylighting. https://doi.org/10.15627/jd.2021.21.
Nourkojouri, H., Shafavi, N., Tahsildoost, M., & Zomorodian, Z. (2021). Development of a Machine-Learning Framework for Overall Daylight and Visual Comfort Assessment in Early Design Stages. Journal of Daylighting. https://doi.org/10.15627/jd.2021.21.
Ghritli, R., Okano, Y., Inatomi, Y., & Dost, S. (2022). Control of growth interface shape during InGaSb growth by vertical gradient freezing under microgravity, and optimization using machine learning. Japanese Journal of Applied Physics, 61. https://doi.org/10.35848/1347-4065/ac99c2.
Zhao, R., Wang, X., Wei, Y., He, X., & Xu, H. (2024). Machine Learning Applied to Electron Beam Lithography to Accelerate Process Optimization of a Contact Hole Layer.. ACS applied materials & interfaces. https://doi.org/10.1021/acsami.3c18889.
Gupta, V., Solomou, A., Limaye, P., Becker, G., Abinesh, M., Meier, H., Valiveti, D., Sun, H., Amalokwu, K., Crawford, B., Manchanda, R., & Kulkarni, K. (2023). A Machine Learning Based Proxy Model for the Rapid Prediction of Hydraulic Fractures. All Days. https://doi.org/10.56952/arma-2023-0315.
Mahmood, M., Khraisheh, M., Popescu, A., & Liou, F. (2024). Processing windows for Al-357 by LPBF process: a novel framework integrating FEM simulation and machine learning with empirical testing. Rapid Prototyping Journal. https://doi.org/10.1108/rpj-01-2024-0057.
Mohammed, A., Tariq, M. A. U. R., Ng, A. W. M., Zaheer, Z., Sadeq, S., Mohammed, M., & Mehdizadeh-Rad, H. (2022). Reducing the Cooling Loads of Buildings Using Shading Devices: A Case Study in Darwin. Sustainability, 14(7), 3775. https://doi.org/10.3390/su14073775
Jakubiec, J. A., & Reinhart, C. F. (2012). The 'adaptive zone' – A concept for assessing discomfort glare throughout daylit spaces. Lighting Research & Technology, 44(2), 149-170.
Najafi, Ahmad. Rahravi-Poodeh, Sanaz. Tadayon, Bahareh. (2023). Optimizing Window Design for Sustainable Daylighting in the Residential Buildings of Isfahan. Contemporary Urban and Architectural Heritage Research Quarterly. 1(1). 131-146
Papadopoulos, S., Woon, W.L., Azar, E. (2018). Machine Learning as Surrogate to Building Performance Simulation: A Building Design Optimization Application. In: Woon, W., Aung, Z., Catalina Feliú, A., Madnick, S. (eds) Data Analytics for Renewable Energy Integration. Technologies, Systems and Society. DARE 2018. Lecture Notes in Computer Science(), vol 11325. Springer, Cham. https://doi.org/10.1007/978-3-030-04303-2_7
Reinhart, C. F., & Walkenhorst, O. (2001). Validation of dynamic RADIANCE-based daylight simulations for a test office with external blinds. Energy and Buildings, 33(7), 683-697.
Al-Tamimi, N. A. M., & Fadzil, S. F. S. (2011). The potential of shading devices for temperature reduction in high-rise residential buildings in the tropics. Procedia Engineering, 21, 273-282.
Tzempelikos, A., & Athienitis, A. K. (2007). The impact of shading design and control on building cooling and lighting demand. Solar Energy, 81(3), 369-382.