- 规划与建设
- 文章编号：1009-6000(2025)10-0034-05
- 中图分类号：TU984 文献标识码：A
- Doi：10.3969/j.issn.1009-6000.2025.10.006
- 项目基金：国家自然科学基金项目（42130402）；深圳市科技计划资助项目（JCYJ202208 18100810024、KQTD20221101093604016、KJZD20230923114911022）。
- 作者简介：赵鹏军，北京大学城市与环境学院，北京大学深圳研究生院城市规划与设计学院；
侯勇企，北京大学城市与环境学院；
陈睿，北京大学深圳研究生院城市规划与设计学院。
- 城市复杂系统空间演进模拟方法研究进展
- Research Progress in Simulation Methods for the Spatial Evolution of Urban Complex Systems
- 赵鹏军 侯勇企 陈睿
- ZHAO Pengjun HOU Yongqi CHEN Rui
- 摘要：
城市是一个有机生命体，其内部蕴含的社会、经济、自然等多元要素，彼此相互嵌套、协同联动，构筑起高度复杂的巨型系统。城市复杂系统具有自组织、自适应、动态开放等特征，在空间上表现为城市要素、功能、结构、特征的不断演进。文章通过文献总结和梳理发现，当前城市复杂系统空间演进量化模拟在技术层面存在5类主流方法，包括描述型方法、动力学方法、交互模拟方法、系统耦合建模方法、知识驱动建模方法等。当前方法呈现从单一时空尺度向多粒度演化分析、从现象描述向机理解释、从局部模型向系统模型的演进趋势。未来，在城市系统的复杂转型新趋势和大模型、人工智能等新技术的推动下，城市复杂系统空间演进研究亟须在 多源异构数据的知识化表征学习、人机协同的系统模型开发等方向取得突破，以实现城市多目标协同优化和全生命周期治理。 - 关键词：
城市复杂系统；定量模拟；空间演进；系统模型；人工智能 - Abstract: A city is an organic living entity whose internal social, economic, natural, and other diverse elements are nested within each other, interacting synergistically to form a highly complex mega-system. Urban complex systems are characterized by self-organization, self-adaptation, and dynamic openness, and spatially manifest through the continuous evolution of urban elements, functions, structures, and characteristics. A review of the literature shows that current quantitative simulation approaches to the spatial evolution of urban complex systems can be broadly classified into five mainstream categories: descriptive methods, dynamic methods, interactive simulation methods, system-coupling modeling methods, and knowledge-driven modeling methods. These approaches are trending from single spatio-temporal scale analyses toward multi-granularity evolutionary analyses, from mere phenomenon description toward mechanism interpretation, and from localized models toward systemic models. Looking ahead, under the new trends of complex urban system transformation and with the support of emerging technologies such as large models and artificial intelligence, research on the spatial evolution of urban complex systems urgently needs breakthroughs in knowledge representation learning of multi-source heterogeneous data and the development of human-machine collaborative system models, so as to achieve multi-objective collaborative optimization and full life-cycle governance of cities.
- Key words: urban complex systems; quantitative simulation; spatial evolution; system modeling; artificial intelligence