الگوریتم سی‌گَن در تولید نقشه حرارتی جانمایی فضایی در طراحی معماری

نوع مقاله : مقاله پژوهشی

نویسندگان

1 استادیار گروه معماری، دانشکده معماری و شهرسازی، دانشگاه علم و صنعت ایران، تهران، ایران.

2 استاد گروه معماری، دانشکده هنر و معماری، دانشگاه تربیت مدرس، تهران، ایران.

3 استاد گروه معماری، دانشکده هنر و معماری، دانشگاه تربیت مدرس، تهران، ایران

4 استاد گروه مکاترونیک، دانشکده مهندسی مکانیک، دانشگاه علم و صنعت ایران، تهران، ایران.

چکیده

نقشه‌های جانمایی فضایی همواره به‌عنوان یکی از اولین مراحل فرآیند طراحی معماری مورد توجه معماران بوده است. چارچوب نظریه معماری سرآمد تأکید دارد که ساختار توپولوژیک و هندسی این نقشه‌ها برگرفته از مفاهیم پنهانی است که خود تحت تأثیر متغیرهای عینی و ذهنی شکل گرفته‌اند. بر اساس فرض پژوهش، نقشه‌های جانمایی فضایی تابع الگوهای نهانی هستند که مبانی شکل‌گیری آن‌هاست. استفاده از قدرت محاسباتی رایانه‮ها برای کمک در پیش‌بینی چیدمان جانمایی فضاها همواره از موضوعات چالش برانگیز در معماری معاصر و چشم‌اندازی برای معماری آینده بوده است. در این مقاله، از روش‌های داده‌محور هوش مصنوعی برای تولید نقشه‌های حرارتی جانمایی فضایی استفاده شده است. روش طراحی شده، برخلاف روش‌های متداول که سعی در تعریف پلان‌های جانمایی بر اساس روابط ریاضیاتی محض دارند، سعی دارد تا با رویکردی طراحی‮مبنا، تابع تولید چیدمان فضایی را برگرفته از تجربه طراحی الگوهای موفق قرار دهد. در این راستا یک مجموعه سیصد پلان جانمایی آپارتمان‌های تهران جمع‌آوری شده است. پلان‌ها در ابعاد مساحتی متفاوت و همه منطبق بر ضوابط طراحی انتخاب شده‌اند. سپس نقشه‌های حرارتی جانمایی این پلان‌ها تهیه شده و چهار نوع تصویر ورودی مختلف برای آموزش مدل هوش مصنوعی تهیه شده است. در این پژوهش از الگوریتم سی‌گَن به‌عنوان یکی از کارآمدترین الگوریتم‌های مولد هوش مصنوعی استفاده، و بر اساس الگوهای جانمایی تهیه شده آموزش داده شده است. این الگوریتم توانایی تنظیم تابع نگاشت، جهت تولید تصویر هدف بر اساس تصویر ورودی را دارد. پس از تکمیل فرآیند آموزش مدل هوش مصنوعی، نقشه‌های حرارتی چیدمان فضایی ده آپارتمان جدید توسط هوش مصنوعی تولید شده و کیفیت جواب‌های پیش‌بینی شده بر اساس پنج ضابطه از پیش تعریف شده ارزیابی شده است. الگوی پیشنهادی پژوهش با انگاره رویکرد طراحی مبنا به فناوری‌های نوین ساختمانی مانند: کاربرد ابرداده‌ها، یادگیری عمیق، یادگیری ماشینی، بهره‌وری و مصرف هوشمندانه انرژی، و انرژی- دیدکارایی هماهنگی دارد.

کلیدواژه‌ها


عنوان مقاله [English]

Training cGan Algorithm for Generating Architectural Layout Heat Map

نویسندگان [English]

  • Morteza Rahbar 1
  • Mohammadjavad Mahdavinejad 2
  • mohammadreza Bemanian 3
  • Amirhossein Davaei Markazi 4
1 Assistant Professor of Architecture, School of Architecture and Environmental Design, Iran University of Science and Technology, Tehran, Iran.
2 Professor of Architecture, Faculty of Art and Architecture, Tarbiat Modares University, Tehran, Iran
3 Professor of Architecture, Faculty of Art and Architecture, Tarbiat Modares University, Tehran, Iran.
4 Professor of Control & Mechatronics, School of Mechanical Engineering, Iran University of Science and Technology, Tehran, Iran.
چکیده [English]

The maps of the space layout have been considered by the architects as one of the first steps of the architectural design process. The theoretical framework of the high-performance architecture emphasizes that the topological and geometrical structure of these maps is adopted from the latent concepts. These concepts were formed under the influence of the subjective and objective variables. According to the research hypothesis, the space layout maps are subject to the latent patterns that are the basis for their formation. Using the computational strength for contributing to predicting the space layouts has always been a controversial issue in contemporary architecture and has been the prospect for future architecture. The current paper used the data-driven artificial intelligence methods for generating the heat maps of the space layout. Despite the conventional methods that try to define the layout plans based on the absolute mathematical relations, the designed method tries to take the spatial layout generator function from the experience of designing successful patterns with a designed based approach. Therefore, a set of 300 plans of the apartments in Tehran has been provided, and four types of different inputs have been supplied for training the artificial intelligence model. In the present research, cGan algorithm was used as one of the most efficient algorithms. This algorithm creates artificial intelligence and has been trained based on the provided layout patterns. This algorithm can regulate the mapping function to generate the target image based on the input image. After completing the process of training the cGAN model, the heat maps of the space layouts of 10 new apartments were tested. Also, the quality of the predicted answers was evaluated based on the predetermined five regulations. The suggested model based on the design-based approach is following modern construction technologies, such as the application of metadata, deep learning, machine learning, efficiency and smart consumption of energy, and energy-view optimization.

کلیدواژه‌ها [English]

  • Artificial Intelligence
  • High-Performance Architecture
  • Data-Oriented Design
  • Future Architecture
Ansarimanesh, M., Nasrollahi, N., & Mahdavinejad, M. (2019). Determination of the Optimal Orientation in the Cold Climate Administrative Buildings; Case Study: Kermanshah”. Armanshahr Architecture & Urban Development, 12(27), 1-9. https://doi.org/10.22034/AAUD.2019.92428
Bassett, H. (2020). Space Syntax in Archaeology & Architectural History. Insightlab of University of Virginia, Available from: http://web.arch.virginia.edu/insightlab/student.php?postid=68
Caetano, I., & Leitão, A. (2020). Architecture Meets Computation: An Overview of the Evolution of Computational Design Approaches in Architecture. Architectural Science Review, 63(2), 165-174. https://doi.org/10.1080/00038628.2019.1680524
Dousti, F., Varij Kazemi, A., & Behzadfar, M. (2018). A New Reading of Sociable Public Spaces: The Nexus between Urban Design and Microsociology. Armanshahr Architecture & Urban Development, 11(22), 39-49.
Fallahtafti, R., & Mahdavinejad, M. (2015). Optimisation of Building Shape and Orientation for Better Energy Efficient Architecture. International Journal of Energy Sector Management, 9(4), 593-618. https://doi.org/10.1108/IJESM-09-2014-0001
Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., & Bengio, Y. (2014). Generative Adversarial Nets, Paper presented at the Advances in neural Information Processing Systems. https://doi.org/10.5555/2969033.2969125
Hadianpour, M., Mahdavinejad, M., Bemanian, M., & Nasrollahi, F. (2018). Seasonal Differences of Subjective Thermal Sensation and Neutral Temperature in an Outdoor Shaded Space in Tehran, Iran. Sustainable Cities and Society, 39, 751-64. https://doi.org/10.1016/j.scs.2018.03.003
Hadianpour, M., Mahdavinejad, M., Bemanian, M., Haghshenas, M., & Kordjamshidi, M. (2019). Effects of Windward and Leeward Wind Directions on Outdoor Thermal and Wind Sensation in Tehran. Building and Environment. (150),164-180. https://doi.org/10.1016/j.buildenv.2018.12.053
Hajian, M., Alitajer, S., & Mahdavinejad, M. (2020). The Influence of Courtyard on the Formation of Iranian Traditional Houses Configuration in Kashan. Armanshahr Architecture & Urban Development, 13(30), 43-55. DOI: 10.22034/aaud.2020.133667.1554
Herthogs, P., Debacker, W., Tunçer, B., De Weerdt, Y., & De Temmerman, N. (2019). Quantifying the Generality and Adaptability of Building Layouts Using Weighted Graphs: the SAGA Method. Buildings, 9(4), 92. https://doi.org/10.3390/buildings9040092
Isola, P., Zhu, J.Y., Zhou, T., & Efros, A.A. (2017). Image-to-Image Translation with Conditional Adversarial Networks. Arxiv Preprint. https://doi.org/10.1109/CVPR.2017.632
Javanroodi, K., Nik, V.M., & Mahdavinejad, M. (2019). A Novel Design-Based Optimization Framework for Enhancing the Energy Efficiency of High-Rise Office Buildings in Urban Areas. Sustainable Cities and Society, 49, 101597. https://doi.org/10.1016/j.scs.2019.101597
Javidmehr, M., & Hashempour, P. (2019). The Influence of Geometry on the Vitality of Architecture, Case Study: Cultural Centers. Armanshahr Architecture & Urban Development, 11(25), 13-26. Available from: http://www.armanshahrjournal.com/article_85043_423f8af188b804edf09faf27929f04ef.pdf
Mahdavinejad, M. (2014). Dilemma of Prosperity and Technology in Contemporary Architecture of Developing Countries. Naqshejahan - Basic studies and New Technologies of Architecture and Planning, 3(2), 35-42. [Persian] Available from: http://journals.modares.ac.ir/browse.php?a_code=A-10-1000-8748&slc_lang=fa&sid=2
Mahdavinejad, M. (2017). High-Performance Architecture: Search for Future Legacy in Contemporary Iranian Architecture. Armanshahr Architecture & Urban Development, 9(17), 129-138. Available from: http://www.armanshahrjournal.com/article_44611_955a20b5cfd1f32308e627ddc8528b91.pdf
Mahdavinejad, M. (2020). Designerly Approach to Energy Efficiency in High-Performance Architecture Theory. Naqshejahan-Basic studies and New Technologies of Architecture and Planning, 10(2), 75-83. Available from: http://journals.modares.ac.ir/article-2-41547-fa.htm
Mahdavinejad, M., & Hosseini, S.A. (2019). Data mining and Content Analysis of the Jury Citations of the Pritzker Architecture Prize (1977–2017). Journal of Architecture and Urbanism, 43(1), 71-90. https://doi.org/10.3846/jau.2019.5209
Mahdavinejad, M., & Javanroodi, K. (2014). Efficient Roof Shapes through Wind Flow and Indoor Temperature, Case Studies: Flat Roofs and Domed Roofs. Armanshahr Architecture & Urban Development, 7(12), 55-68. Available from: http://www.armanshahrjournal.com/article_33518_f85664c41afa683f7c7a2cd740478fbe.pdf
Mahdavinejad, M., & Refalian, G. (2011). Parametric Algorithms for Unity of Architecture and Construction. Scientific-Research Journal of Iranian Scientific Association of Architecture & Environmental Design, 2(2), 61-67. DOI: 10.30475/ISAU.2011.61940
Mahdavinejad, M., & Shahri, S. (2015). Contemporization of Tehran Traditional Architecture by Parametric Algorithm. Hoviatshahr, 8(20),31-44. Available from: http://hoviatshahr.srbiau.ac.ir/article_6419_697f157aebfd2e62eeb7b8574b770aeb.pdf
Mahdavinejad, M., Zia, A., Larki, A.N., Ghanavati, S., & Elmi, N. (2014). Dilemma of Green and Pseudo Green Architecture Based on LEED Norms in Case of Developing Countries. International Journal of Sustainable Built Environment, 3(2), 235-246. https://doi.org/10.1016/j.ijsbe.2014.06.003
Nisztuk, M., & Myszkowski, P.B. (2019). Hybrid Evolutionary Algorithm Applied to Automated Floor Plan Generation. International Journal of Architectural Computing, 17(3), 260-283.
Pilechiha, P., Mahdavinejad, M., Rahimian, F.P., Carnemolla, P., & Seyedzadeh, S. (2020). Multi-Objective Optimisation Framework for Designing Office Windows: Quality of View, Daylight and Energy Efficiency. Applied Energy. 261, 114356. https://doi.org/10.1016/j.apenergy.2019.114356
Pramanik, P.K.D., Mukherjee, B., Pal, S., Pal, T., & Singh, S.P. (2020). Green Smart Building: Requisites, Architecture, Challenges, and Use Cases. In Green Building Management and Smart Automation (1-50). IGI Global. https://doi.org/10.4018/978-1-5225-9754-4.ch001
Rahbar, M., Mahdavinejad, M., Bemanian, M., & Davaie Markazi, A.H. (2020). Predicting Environmental Sustainable Parameters with Applying Artificial Neural Network. Journal of Sustainable Architecture and Urban Design, 7(2), 169-182. https://doi.org/10.22061/jsaud.2019.4501.1333
Rahbar, M., Mahdavinejad, M., Bemanian, M., Davaie Markazi, A.H., & Hovestadt, L. (2019). Generating Synthetic Space Allocation Probability Layouts Based on Trained Conditional-GANs. Applied Artificial Intelligence, 33(8), 689-705. https://doi.org/10.1080/08839514.2019.1592919
Saadatjoo, P., Mahdavinejad, M., & Zhang, G. (2018). A Study on Terraced Apartments and Their Natural Ventilation Performance in Hot and Humid Regions. Building Simulation, 11(2), 359-372. https://doi.org/10.1007/s12273-017-0407-7
Talaei, M., & Mahdavinejad, M. (2019). Probable Cause of Damage to the Panel of Microalgae Bioreactor Building Façade: Hypothetical Evaluation. Engineering Failure Analysis, 101,9-21. https://doi.org/10.1016/j.engfailanal.2019.02.060
Talaei, M., Mahdavinejad, M., & Azari, R. (2020). Thermal and Energy Performance of Algae Bioreactive Façades: A Review. Journal of Building Engineering, 28, 101011. https://doi.org/10.1016/j.jobe.2019.101011
Zhu, J.Y., Park, T., Isola, P., & Efros, A.A. (2017). Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks. In Proceedings of the IEEE International Conference on Computer Vision, 2223-2232. https://doi.org/10.1109/ICCV.2017.244
Ziaee, A., Moztarzadeh, H., & Movahec, K. (2020). The Role of Parametric System in the Analysis of Sim Van Der Ryn’s Ecological Architecture Principles in Iranian Plateau. Armanshahr Architecture & Urban Development, 13(30), 167-182. https://doi.org/10.22034/aaud.2020.183579.1862