【摘要】 超大城市人员、财富与关键资源高度集聚，在多灾种耦合与系统复杂性持续增强的背景下，传统以静态风险区划为主的防灾模式已难以支撑精细化与动态化治理需求。系统梳理城市防灾理念与技术演进，对于构建面向未来的韧性治理框架具有重要意义。通过回顾城市防灾地图，从第一代单一危险性区划的静态地图、第二代综合暴露度与易损性的风险地图，到融合多源数据、数字孪生与人工智能的第三代动态韧性地图的发展脉络。在此基础上，面向区域尺度广与系统耦合复杂等挑战，提出以“智模—智感—智算—智判—智控”为主线的城市韧性防灾智能体系，涵盖跨系统建模、多源感知互联、建筑群与基础设施灾变响应计算、区域韧性评估及韧性地图运维调控。构建了城市防灾动态韧性地图框架，实现城市运行状态的可感知、可预测与可调控，并在灾前评估、灾时/灾后监控决策及城市更新等场景中验证其可行性。该体系通过跨系统建模与多源智能融合，为超大城市防灾治理提供从风险识别到运行调控的整体技术支撑。未来需在数据治理、AI与物理机理融合及制度协同方面持续推进，推动韧性地图向城市长效治理基础设施演进。更多还原

【Abstract】 Megacities concentrate vast populations, wealth, and critical resources, making the enhancement of disaster prevention, mitigation, and rapid recovery capacities a cornerstone of sustainable urban development. Against the background of increasingly frequent compound and cascading disasters, traditional approaches to urban disaster prevention are facing new challenges in terms of regional scale, system coupling, and resilience management. This paper first reviews the evolution of urban disaster prevention concepts and disaster-mapping techniques. The development trajectory is summarized from the first generation of static maps that primarily focused on single-hazard zoning, to the second generation of disaster risk maps that explicitly incorporate the exposure and vulnerability of affected elements, and further to the third generation of dynamic resilience maps for urban disaster prevention that integrate multi-source heterogeneous data, digital-twin technologies, real-time sensing, and artificial intelligence. This progression reveals a clear shift in disaster-risk characterization from static description to dynamic assessment, from single indicators to multidimensional integration, and from hazard-centered thinking to resilience-oriented governance. Building upon this understanding, and in response to contemporary challenges such as large regional coverage, strong interactions among urban subsystems, and high uncertainty in disaster processes, this paper proposes a technical pathway for constructing an intelligent governance system for disaster-resilient cities. The proposed pathway includes:(1) multi-granularity, cross-system digital-twin modeling of urban spaces and infrastructure systems;(2) ubiquitous perception and full-domain interconnection of multi-source urban data;(3) intelligent computation and rapid prediction of disaster-induced responses of building clusters and lifeline infrastructure;(4) urban-scale resilience assessment based on functional performance and recovery processes; and(5) digital operation, visualization, and regulation of resilience maps to support refined governance. Furthermore, typical application scenarios of the proposed framework are explored, including disaster preparedness and risk prevention prior to events, real-time monitoring, early warning, and decision support during and after disasters, as well as its role in urban renewal and resilient urban planning. These applications demonstrate the broad prospects of the proposed technical pathway in enhancing the disaster prevention, mitigation, and recovery capabilities of megacities, and in supporting the construction of safe, resilient, and intelligent urban governance systems.更多还原