The Application of Evolutionary Algorithms and Shape Grammar in the Design Process Based upon Traditional Structures

Abstract

Problem statement: Although the application of shape-based methods in the generative design systems has been studied as a means for describing or analyzing a style, regional architecture, or a specific design in numerous works of literature, the complexity and multiplicity of the structural grammars, as well as the lack of flexibility in the system for evaluation and optimization of new models has somewhat dwindled the applicability of these approaches. However, in addition to investigating and extracting the shape grammar of a traditional structure, the potential of generating novel designs based upon the shape rules indicates the necessity of further examination of the shape-based approaches. Research objectives: Addressing the issue, this study proposes a hybrid generative evolutionary approach. The use of generative evolutionary methods can provide fresh insight into the process of analyzing traditional structures and designing new ones based on these structures.Research method: The methodology is based upon library studies with the analytical-descriptive approach to examine the related literature review. Furthermore, with the consideration of Kashan’s traditional houses as a case study, it applies the combination of shape grammar method for analyzing configuration characteristics and physical-spatial relationships in historic houses, with genetic algorithm for optimizing the new generated forms. Conclusion: The results of the study indicate the effectiveness of generative evolutionary methods in improving and accelerating the process of analyzing the characteristics of traditional structures as well as producing a variety of new designs that can preserve the spatial essence of traditional structures.