【作者】 陈伟；

【导师】 翟国方；

【作者基本信息】 南京大学 ， 城市与区域规划， 2019， 博士

【摘要】 应急避难场所建设对构建完善的城市防灾设施网络至关重要。当前,市域或市区尺度下的避难场所规划是我国防灾规划实践的主要方向,其内容与成果形式属总体规划层次范畴,偏指导性,且普遍存在诸如避难需求分析不合理、场所布局缺乏量化依据、对紧急避难场所关注不够、无法对城市规划形成有效反馈等问题,导致规划实施效果不理想,避难场所建设缓慢。虽有部分学者注意到了相关问题并进行了研究,但目前尚未形成较好的解决方案。此外,大数据背景下的城市规划技术方法正处于持续创新的过程之中。基于这些事实,本研究从多尺度规划视角出发,结合多源数据,建立了不同空间尺度下避难场所规划的主要内容与关键问题研究方法框架,并分别针对市域、区级、片区三个案例区域展开了实证。全文共分为四个部分:第一部分为绪论,具体阐明本研究的源起、目标、意义、内容、方法和技术路线。第二部分为理论研究基础与方法框架,包括第二章和第三章的内容。第二章对相关研究领域、实践进展进行了综述,分析对比了国内外研究的主题方向并发现了存在的不足之处,同时对我国避难场所规划实践中的主要方法性问题进行了总结,为本研究的创新性奠定基础。第三章从研究的“尺度”出发,构建了避难场所多尺度规划的主要内容与关键问题研究方法框架,包括数据选择、模型的构建思路与实现技术流程等,用以指导第三部分的实证探索。第三部分为实证研究,基于第二部分已构建的内容与方法框架,结合不同空间尺度的案例区域展开避难场所规划方法实证研究,包括第四、五、六章,是本文的主体内容。市域尺度下的避难场所规划方法实证以广州市为例。研究对象为包括中心避难场所在内的中长期固定避难场所,以满足夜间灾害情景下的避灾疏散需求为主要目标。首先对避难场所可用资源进行调查与筛选,结合危险性、脆弱性、抗灾能力相关因子分析了地震、洪涝、风暴潮、地质灾害的风险,以了解各区不同灾害的风险等级,为避难场所规划提供基本依据与方向引导;同时,为确保场址选择的安全可靠,建立评价指标体系对多灾环境下的避难场所空间适宜性进行了评估;最后,利用灾害潜在影响区域和规划人口密度的叠加分析计算了应急避难需求,据此对避难场所布局展开了研究。区级尺度下的避难场所规划方法实证以广州市荔湾区、越秀区、天河区和海珠区为例。研究对象主要为短期固定避难场所,避难场所防灾目标由市域规划尺度的夜间防灾需要转向满足昼夜灾害情景下的疏散需求。首先,基于POI数据、带户数属性的居住建筑数据以及行业从业人员统计数据分析了城市人口在白天和夜间的分布,并以此预测了应急避难需求;然后,通过Python程序语言编程设计了人员完全疏散模型和场所容量制约模型,考察了避难场所的供需状态,为场所空间优化提供量化依据;最后,利用空间优化模型实现需求点剩余未疏散人员向周边目标地块的再分配,由此提出了避难场所空间优化面积指标与场所类型建议。片区尺度下的避难场所规划方法实证以天津市中心城区的“小白楼CBD”片区为例。研究对象为紧急避难场所,重点探讨了手机信令数据在规划中应用实现的技术方法。在人口动态变化的背景下,首先通过手机信令数据掌握了较高空间精度的城市人口时空分布,发现了人口活动的热点地块及热点时间,为避难场所规划提供了方向性引导。同时,为预测应急避难需求,以Python语言编程设计了基于手机信令数据的人员分析模型对驻留人员与流动人员进行属性识别,并估算了地块的合理避难人员数量。据此,结合第五章提出的两种疏散模拟模型和避难场所空间优化模型对区域内避难场所供需状态进行了分析,提出了目标地块避难场所空间优化的面积规模建议。进一步地,运用ArcGIS模型构建器(Model Builder)构建了任意两点的疏散路径可视化模型,并通过编程实现了模型的自动处理,批量可视化了每个需求点至其周边若干避难场所的疏散路径。第四部分为本研究的主要结论、创新点及未来的研究方向。主要结论有:1)构建的避难场所多尺度规划方法体系经实证检验具有一定的科学性与可操作性,可从多层次规划延续的角度为避难场所建设提供较强的规划支撑,利于促进规划的有效实施。2)多源数据的使用能够实现不同规划尺度下避难需求的逐层细化分析,为避难场所供需评估提供了量化依据,从而引导不同类型避难场所的布局研究。3)综合运用跨平台技术提出了避难场所规划实践与研究中关键难点的应对方法,能够为大数据应用趋广背景下的防灾规划乃至城市规划领域的问题研究提供方法性参考。研究在避难场所多尺度规划体系及其主要内容与关键问题研究方法框架的构建、大数据在避难场所规划研究中的应用、避难场所规划对城市规划的量化反馈三方面取得了一定的创新。更多还原

【Abstract】 The construction of emergency shelter is essential for building a sound network of urban disaster prevention facilities.At present,planning at the city or urban scales is the main direction in disaster prevention planning practice in China.From the perspective of content and results,it belongs to the master planning category,and there are widespread problems,which are unreasonable analysis of shelter need,lack of quantitative basis for spatial layout of emergency shelter,insufficient attention to temporary shelter,and inability to form effective feedback on urban planning,etc.As a result,it has led to unsatisfactory planning implementation and slow construction of shelters.Although some scholars have noticed related problems and conducted studies,no better solutions have been formed yet.In addition,urban planning techniques in the context of big data era are in the process of continuous innovation.Therefore,this study establishes the main content and key problem research method framework of shelter planning under different spatial scales from the perspective of multi-scale planning,combined with multi-source data,and conducted empirical study of three cases at the city,borough and district scales.The dissertation is structured into four parts:The first part,introduction,clarifies the origin,goal,significance,content,method and technical route of this study.The second part is the theoretical research basis and method framework,including the contents of Chapters 2 and 3.Chapter 2 reviews the relevant research fields and practical progress,analyzes and contrasts the thematic research directions home and abroad and identifies the existing deficiencies.Then,it summarizes the main methodological problems in the shelter planning practice in China,which lays a foundation for this study.The third chapter starts from the concept of scale,which is related to the structure of main body and key problem to solve in the research method framework of multi-scale planning of emergency shelter.It includes data selection,model construction ideas and technology flow,etc.,to guide the third part of the empirical study.The third part,the empirical study,is developed on the basis of constructed content and method framework,including Chapter 4,5,and 6 which constitute the main content of this dissertation.The empirical study on planning method at the city scale takes Guangzhou as a case.The research objects are medium-and long-term fixed shelters including central ones,and the main goal is to meet the needs of disaster avoidance under nighttime disaster scenarios.Firstly,the resources available for the shelters are investigated and screened.Combining hazard,vulnerability and order to ensure the safety and reliability of site selection,an index system is established to assess the spatial suitability of shelters in a multi-hazard environment.Finally,the emergency shelter need is calculated using the superposition analysis of the potential impact areas of disasters and the planned population density,and the layout of the shelters is studied accordingly.The empirical study on planning method at the borough scale takes Liwan District,Yuexiu District,Tianhe District and Haizhu District in Guangzhou as cases.The research objects are mainly short-term fixed shelters,and the disaster prevention targets of the shelters are shifted from the nighttime disaster prevention needs at the city planning scale to the evacuation requirements under the day and night disaster scenarios.Firstly,based on the POI data,the residential building data with the number of households and the statistics of the industry’s employees,the distribution of urban population during the day and night is analyzed,and the emergency shelter need is predicted.Then,two evacuation simulation models are designed by Python programming language to investigate the supply and demand status of the shelters,which provides a quantitative basis for layout optimization.Finally,a spatial optimization model is used to realize the redistribution of the remaining nonevacuated people from the demand point to the surrounding target plots,and the suggestions on the spatial optimization area index and type of shelter are put forward.The empirical study on planning method at the district scale takes "Xiaobailou CBD" in the downtown area of Tianjin as a case.The research objects are temporary emergency shelters,with a focus on the technical methods of the application of mobile signaling data in planning.In the context of demographic dynamics,firstly,the spatial and temporal distribution of urban population with higher spatial precision is grasped by mobile phone signaling data,and hot spot areas and hotspots of population activities are found,which provided directional guidance for shelter planning.At the same time,in order to predict the emergency shelter need,the personnel analysis model based on mobile phone signaling data is designed in Python language to identify the resident and mobile personnel,and the number of reasonable potential refugees in the district is estimated.Based on this,combined with the two evacuation simulation models and the spatial optimization model proposed in Chapter 5,the supply and demand status of the shelters in the district is analyzed,suggestions on optimizing the area and scale of shelters in target plots are put forward.Furthermore,the ArcGIS Model Builder is used to construct the two-point evacuation route visualization model,and the automatic processing of the model is realized by programming.The evacuation routes from each demand point to the shelters around it are displayed in batches.The fourth part concludes this dissertation,and identifies its innovations and directions of future works.The main conclusions are as follows:1)The constructed multi-scale planning method system of emergency shelter has certain scientific and operability through empirical test,which can provide strong planning support for the construction of shelters from the perspective of multi-level planning continuation,and is conducive to promoting the effective implementation of planning.2)The use of multi-source data can realize the layer-by-layer refinement analysis of shelter need under different planning scales,and provide a quantitative basis for the supply and demand assessment of shelters,thus guiding the layout of different types of shelters.3)Comprehensive application of cross-platform technology is proposed to solve the key difficulties in the practices and studies of shelters.It also can provide a methodological reference for disaster prevention planning and urban planning in the context of big data application.Finally,this paper has made some innovations in the scholarship of multi-scale planning framework of shelter system,the application of big data in the urban shelter system planning,and the quantitative responses of shelter planning on urban planning.更多还原