인공지능(AI)을 활용한 건축디자인 정보처리 모델에 관한 연구

Abstract

Recently, there is a lot of discussion in the architectural world about the impact that artificial intelligence image generation models will have on architectural design. Artificial intelligence uses information processing theory to analyze and model human cognitive abilities. According to information processing theory, architectural design is an information processing activity that establishes reasonable concepts to solve spatial problems and processes visual and conceptual information that corresponds to them. In this study, the five information processing processes of Akin's (1986) information processing model, DIPS (design information processing system), 'Information acquisition', 'Information representation', 'Information projection', 'Information confirmation', and 'Regulation of control'. The characteristics of architectural design thinking types, ‘convergent thinking’, ‘divergent thinking’, and ‘analogical thinking’ were examined. Through this, I identified the characteristics of representative design-related artificial intelligence NLP, GAN, and DNN and proposed an architectural design information processing model using artificial intelligence. The purpose of this paper is to examine the architectural design information processing model and compare the characteristics of the thinking style that occurs in the process with the characteristics of the artificial intelligence model to analyze the possibility of using artificial intelligence in architectural design and to provide architectural design information using artificial intelligence. The goal is to propose a processing model.