大模型融入城市规划专业工作的技术范式与模型构建OACHSSCD

The paper starts by examining the nature of professional work in urban planning and the typology of artificial intelligence.It focuses on the technical para-digm through which large models can be embedded into professional urban planning practice and proposes an approach for constructing professional planning large mod-els.The paper characterizes urban planning professional work as a"diagnosis-infer-ence-treatment"professional working mode,in which domain knowledge is used to provide solutions.By distinguishing between perceptual intelligence and cognitive in-telligence,the study demonstrates that large models are capable of supporting this professional working mode.Building on this understanding,a technical framework for constructing professional planning large models is proposed,emphasizing domain knowledge acquisition through two stages:diagnostic training and treatment training.An empirical case is presented to demonstrate that professional planning large mod-els can acquire professional planning capabilities,including the ability to identify spa-tial problems,infer underlying causes,and propose planning strategies.The findings indicate that the technical paradigm for embedding large models into urban planning practice consists of professional large models equipped with both general knowledge and domain-specific knowledge,operating within the"diagnosis-inference-treat-ment"working mode to support problem-oriented and law-oriented planning para-digms.Professional planning large models serve as the carrier of this paradigm and are developed through domain-knowledge-centered learning based on general large models,implemented via diagnostic and treatment training.Such domain-specific pro-fessional large models can be effectively embedded into planning practice and func-tion to enhance the professional capacity of planners.