AIGC技术下的建筑生成设计方法初探——以Prompt关键词生成建筑意象的整体设计过程为例  

Exploration on the Generative Architecture Design Method with AIGC Technology:A Case of the Overall Design Process of Generating Architectural Image with Prompt as a Key Word

作　　者：黄龑枫 周毅荣 Huang Yanfeng;Zhou Yirong

机构地区：[1]同济大学建筑与城市规划学院

出　　处：《城市建筑》2023年第15期202-206,213,共6页Urbanism and Architecture

摘　　要：近年来,随着大型语言模型(LLM)和各类生成式神经网络技术的突破,建筑生成设计对规则、算法等基于建筑学本体建构的研究脉络不断完善,但是从建筑师设计流程和工作环节中交互的具体文本和出图环节方面讨论建筑生成设计的研究较少。本研究以Prompt关键词生成建筑意象的整体设计过程为例,通过文献研究与实证调研方法,从AIGC现有的技术出发,总结目前AIGC的三个生成路径方法,而后提出了一个建筑方案设计过程全生成的优化方法设想。In recent years,with the breakthroughs in neural network technology such as large language models and various generative models,the study on generative building design through the ontology of architecture,rules,algorithms,etc.,has been continuously improved.However,there has been little research on generative building design based on the specific text and graphics interactions within the architect’s design process and workflow.This paper takes the overall design process of generating architectural image with Prompt as a key word as an example.Through literature research and empirical research methods,it summarizes the current three generative pathways of AIGC.Then,an optimized method for the whole generative process of architectural scheme design based on AIGC’s existing technology is proposed.

关 键 词：AIGC 建筑生成式设计 建筑方案设计流程 建筑意象 

分 类 号：TU201[建筑科学—建筑设计及理论]