随着我国城市化水平的提高，原本相对独立的城市个体开始整合形成城市群网络。尤其是我国重点推动的粤港澳大湾区，区内各城市的相互作用在空间发展中扮演越来越重要的角色。对空间扩张过程进行模拟预测，可以为城市规划和生态管理提供决策参考。然而以往研究主要适用于城市个体，较难准确地反映城市群的发展趋势。因此为深入研究城市群空间扩张模拟这一关键问题，本项目拟以粤港澳大湾区为实例开展以下研究：1）构建基于选择性集成学习的城市联系强度挖掘模型，实现多源数据的优势互补；2）结合城市联系强度与城市流，分析城市相互作用对城市群空间扩张的影响机理与贡献程度；3）构建顾及城市相互作用的元胞自动机，从不同区域尺度揭示城市群的空间扩张特征；4）借鉴知名城市群的发展经验，进行城市相互作用驱动下的情景分析以预测未来扩张。本项目研究将有助于深入了解城市群复杂网络的空间发展规律，能为我国城市群的规划与发展提供决策参考和技术支撑。

With the development of urbanization in China, many single cities have been integrated as urban agglomerations. Especially in Guangdong-Hong Kong-Macau Greater Bay Area, inter-city interactions play a growing role in regional development. Simulating and predicting spatio-temporal urban expansion could provide valuable decision support for urban planning and environmental management. However, previous attempts are mainly suitable for a single city, and thus fail to accurately reflect the growth trends for urban agglomerations. Therefore, this project aims to thoroughly investigate urban expansion modeling at urban agglomeration level, and will: (1) develop an inter-city connection measuring model based on selective ensemble learning, which can combine the advantages of multi-source datasets; (2) integrate inter-city connections with urban flows, and then analyze their importance to urban expansion at urban agglomeration level; (3) develop a CA model that considers the spatial influence of inter-city interactions, and then characterize the urban agglomeration development at different regional scales; and (4) carry out scenario analysis to explore various future urban expansion possibilities by referring to the experience from different famous urban agglomerations. This project could enrich our understanding of urban agglomeration development, and could provide valuable guidance and technical support for the planning and development of urban agglomerations.