A MULTI-OBJECTIVE GENETIC ALGORITHM FRAMEWORK FOR EARLIER PHASES OF ARCHITECTURAL DESIGN A Case Study

Kocabay S., Alaçam S.

22nd CAADRIA Annual International Conference on Computer-Aided Architectural Design Research in Asia (CAADRIA), Suzhou, Çin, 5 - 08 Nisan 2017, ss.293-302 identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Basıldığı Şehir: Suzhou
  • Basıldığı Ülke: Çin
  • Sayfa Sayıları: ss.293-302
  • İstanbul Teknik Üniversitesi Adresli: Evet

Özet

This paper presents an algorithmic framework proposal for implementation of a multi-objective genetic algorithm (MOGA) in architectural design process. Different than the previous studies, we introduce a dynamic and extendible modular framework for multiple objectives. The objective modules with different fitness functions are connected simultaneously in the Rhino/Octopus interface, after multiplication with a constant value or a variable. In this study, we discuss the potentials and limitations of MOGA in 3D form generation, implications of MOGA in a case study and the qualitative and quantitative changes in relation to the change of constant value/the impact ratio of competing objectives. The outcomes of the case study are investigated based on its potentiality in providing feedback in the earlier phases of decision processes in design.