| 摘要: |
| 为揭示城市避难场所的运行逻辑，构建空间规划干预原理，在梳理了避难场所发展脉络、主要规划研究内容的基础上，发现避难场所现行规划模式，依赖服务人群覆盖范围最大化、疏散距离最小化等静态指标，构建了千人指标“服务半径”的规划模式，这与当前灾害多灾种化、避难群体化、救援综合化的客观现实需求难以适配。为此，基于灾前-灾时-灾后的时序演进特征，提炼灾害发生-紧急避险-应急救援的避难响应时空演化特征与需求，结合网络科学理论，同时响应“平急结合”理念，提出“服务网络”的规划范式，构建“灾害信息-行为轨迹-设施拓扑”的三层立体交叉网络，从而建立城市避难场所规划的网络化模式逻辑框架，形成“灾害传导与约束→行为模拟与分布→设施关联与耦合”的配置优化思路,为避难场所的系统化规划与建设提供参考依据。 |
| 关键词: 城乡规划 避难场所 服务半径 服务网络 研究进展 |
| DOI： |
| 分类号:TU984.11+6 |
| 基金项目:重庆市自然科学基金：重庆市生命线网络级联故障表达、模拟与空间干预（CSTB2023NSCQ-MSX1003）；国家“十四五”重点研发计划课题：低碳生态乡村社区规划建设整治技术与评价体系研究（2024YFD1600401） |
|
| Research progress on urban emergency shelter planning: from "service radius" to "service network" |
|
huang yong, zhu jia yi, liu zhi cheng
|
|
Chongqing University
|
| Abstract: |
| This study seeks to reveal the underlying operational logic of urban emergency shelter systems and to formulate a robust principle for spatial planning intervention. Our inquiry begins with a comprehensive review of the evolutionary trajectory and dominant research themes within emergency shelter planning. This review critically identifies a persistent reliance on static, location-allocation metrics, such as maximizing population coverage within predefined areas and minimizing geometric evacuation distance, which dominates prevailing planning practices. This conventional approach has crystallized into a planning paradigm predominantly governed by the "service radius" standard per capita, a model fundamentally rooted in facility-centric and geometric efficiency assumptions.
However, this static paradigm demonstrates increasing inadequacy when confronted with the complex, dynamic realities of contemporary urban disaster management. Modern challenges currently facing cities are as follows: (1) multi-hazard scenarios, where sequential or compound disasters demand adaptable rather than fixed responses; (2) collectivized evacuation behaviors, where group dynamics, information dissemination, and social networks critically influence movement patterns; (3) integrated rescue operations, which require seamless coordination between shelters, transportation networks, and medical facilities, a coordination poorly captured by isolated radius-based planning. The existing model, therefore, suffers from a discrepancy between its static spatial framework and the dynamic, interconnected processes of actual disaster response.
To bridge this critical gap, this paper proposes a paradigm shift: from a "service radius" to a "service network" framework. This shift is anchored in two foundational pillars. First, we analyze the spatiotemporal evolution of shelter demand and response, tracing the sequence from disaster occurrence to emergency evacuation and through post-disaster rescue and recovery. This phased analysis (pre-disaster, during-disaster, post-disaster) reveals how shelter needs morph over time, influencing and being influenced by population behavior and system performance. Second, we integrate principles from network science theory, which provides the analytical tools to model relationships, flows, and systemic robustness, moving beyond the analysis of individual shelter points.
The proposed "service network" paradigm materializes as a three-layered, interdependently structured network model:
1. The disaster information layer: This layer models the spatial propagation like flood inundation, seismic shaking, defining the constraints and triggers for evacuation.
2. The behavioral trajectory layer: This layer simulates the movement patterns of evacuees. It incorporates factors beyond distance, such as road capacity, congestion, real-time information, and social ties, generating probabilistic flow networks rather than deterministic assignment.
3. The facility topology layer: This layer represents the physical shelter system not as isolated units but as a network of nodes (shelters) and edges (connecting transportation corridors). Its topology, including connectivity, capacity, and centrality, determines its ability to receive and redistribute flows from the behavioral layer under the constraints of the disaster layer.
The core innovation lies in the dynamic intersection and coupling of these three layers. The disaster layer constrains viable trajectories; the behavioral layer loads the facility network with demand flows; and the facility topology's resilience feeds back into achievable evacuation outcomes. This integration directly embodies the "peacetime-emergency integration" principle, as the network's structure is planned during peacetime but is validated against dynamic emergency scenarios.
Consequently, this research establishes a novel networked logic framework for urban emergency shelter planning. This framework dictates a reconfigured planning sequence: "disaster transmission and constraints → behavioral simulation and distribution → facility correlation and coupling." Optimization no longer seeks merely to place shelters within a radius of demand points but to enhance the overall network's performance metrics—such as its redundancy, throughput, equitable access, and adaptive capacity under disruption.
In conclusion, this study provides a more systematic and adaptive reference for the systematic planning and construction of emergency shelter systems. It advocates for a perspective that views shelters as interdependent components of a larger socio-technical-ecological network, thereby offering a more resilient and actionable pathway to safeguard urban populations in an era of escalating and complex disaster risks. |
| Key words: urban and rural planning emergency shelter service radius service network research progress |
