XVIII. MİMARLIKTA SAYISAL TASARIM ULUSAL SEMPOZYUMU
A Review on the Application of Artificial Intelligence and Neural Fuzzy Logic Method for the Design of Energy-Efficient Buildings
Yayıncı:
Balıkesir Üniversitesi Yayınları
Artificial intelligence in architecture can be applied to various disciplines and used as a decision-making tool. However, its application areas in architecture are expanding with multiple methodologies. Nowadays, the usage areas of these methods, which are also integrated into architecture course contents, are pretty standard, but they are still in the development stage. While there are many alternative applications, especially for generating visuals, it has been examined in the literature that there are fewer applications for methods that require software with digital content. Applying these methods, which have developed as a great alternative to human calculation or overlooking, especially in multi-criteria situations before architectural design, determines the speed and path of decision-making in starting the design. This research provides a comprehensive and in-depth systematic review of recent work on applying artificial intelligence technologies to create energy-efficient buildings. In addition to detailing the principles and applications of AI-based modeling approaches commonly used in predicting building energy use, an evaluation of the work carried out in major AI fields, exceptionally energy efficient building, is made and explained through an example building in terms of architectural design. Within the scope of the study, a case study was carried out by determining the necessary parameters for the indoor pool structure, which is the selected sample building, to provide ideal energy efficiency. Indoor and outdoor temperature values, thermal transmittance value of insulation material, ventilation in the space, presence of a shading element on the facade, building orientation (south, north, east, and west), and window wall ratio are the parameters evaluated as input in this study. In addition, the research outputs include the determination of the targeted annual energy requirement within the scope of the study. Building features with the lowest energy requirements in each group were examined by splitting them into eight groups to compare the numerical data collected. The findings of the research show that the error value ratio found with the neural fuzzy logic ANFIS method is 0.0033174 for the accuracy in the emergence of the values that will guide the building design, and when evaluated by regression analysis, the R² value is 0.9997 (99%). It is anticipated that this study is expected to help architects and contribute to the literature as a method that can generate ideas for new studies to be conducted due to the scarcity of studies on neural fuzzy logic in architecture. It is emphasized that the methods used in this study can significantly improve the energy efficiency and cost-effectiveness of buildings designed to provide a comfortable indoor environment for building occupants. Finally, the study discusses future research on using artificial intelligence in energy-efficient buildings.
Yayınevi Bilgileri
Alıntı Yap
Benzer Yayınlar
Kitap Bölümleri
Bu Sayfayı Yazdırın
Endnote
Mendeley