【作者】 贾婧；

【导师】 冯启民；

【作者基本信息】 中国海洋大学 ， 环境工程， 2008， 博士

【摘要】 城市综合防灾规划的特点体现在多灾种性和多系统性。基于研究与实践的经验,构建了城市综合防灾规划的体系结构,其首要环节即为土地利用综合防灾规划。基于土地利用综合防灾规划的概念、作用、特点等的分析,归纳出这项研究主要包括两大部分内容:第一为结合各个土地单元的单灾种危险性分析结果与土地属性,进行抗该灾种的用地适宜性评价(例如土地抗震适宜性评价);第二为综合各个灾种的危险性分析结果、土地属性以及土地承载体(整个城市系统)的多方面属性,研究土地利用综合灾害风险。在这些研究内容中所处理的数据基本上为空间数据;所涉及的因素有明显的随机性和非线性的特点;并且涵盖的时空范围广、数据量大,是一项十分复杂的非结构化问题。随着空间科技(ST)和信息科技(IT)的快速发展,空间分析数学模型和空间决策支持系统(SDSS)的构建过程更加专业化、模块化、智能化。本文结合土地利用综合防灾规划的理论分析与空间决策支持技术,在以下几个方面进行了研究:(1)着重研究了土地利用综合防灾规划中地震这一灾种,研究目标归纳为建设用地抗震适宜性评价。基于对城市抗震防灾规划、地震灾害预测等相关领域的理论研究和实践经验,构建了建设用地抗震适宜性评价的四个影响指标(场地类别指标、液化指标、地震危险性指标以及地面稳定性指标)的计算模型及其综合评价模型。●场地类别指标的计算模型是基于土层等效剪切波速和覆盖层厚度这两个区域化变量的趋势面拟合和变异函数统计的地统计学分析方法,对离散钻孔点的数据进行空间分布模拟。然后分别通过确定性方法和模糊综合评价方法求得场地类别划分的栅格。以项目数据为算例阐述了上述过程。●液化指数指标的计算模型改进了规范中传统的确定性方法,采用了基于统计模拟方法(蒙特卡洛模拟法)的概率模型计算钻孔点对不同液化等级的概率值。再通过地统计学的分析模型进行空间模拟,得到不同液化等级的概率分布栅格。●在研究地震危险性分析国内外现有方法的基础上,提出了适用于城市抗震防灾规划及综合防灾规划的基于模糊数学理论和空间分析过程的地震危险性分析方法,FSHASA。该方法的基础空间数据库包括了断层、历史地震、地质、布格重力异常等;模型库包括了FSHASA的11个评价指标的空间分析数学模型,以及这些指标的模糊综合评价模型;方法库包括了确定指标权重的层次分析法、函数拟合的麦夸特法等。最后选取6个地区进行了FSHASA法的测试。●以区别于地壳稳定性和地基稳定性,首先界定了地面稳定性的概念,并构建了地面稳定性分析的多层评价指标体系。目标层包括构造、历史地质灾害、人类工程活动、地形地貌这四个指标。并采用多层模糊综合评价的方法分析该多层评价指标体系。在这四个建设用地抗震适宜性评价指标分析的基础上,用层次分析法对这四个指标的结果栅格进行了综合分析,并且考虑到最低因子限制率理论,用修正栅格对综合评价结果进行了修正。通过工程实例对该计算过程进行了说明。(2)土地利用综合防灾规划的第二个研究内容即为土地利用综合灾害风险分析。很多城市在进行综合防灾规划工作之前,已经具备了各个灾种的危险性区划结果,因此需要特别研究的是基于这些结果,综合土地因素、城市环境及其它相关社会因素,分析不同土地区块的潜在灾害风险。综合考虑灾害的致灾因子、孕灾环境、承灾体等因素,构建了计算土地利用综合灾害风险指数的多层指标体系。该体系为3层,目标层包括5个指标(用地功能、直接危险性、次生危险性、承灾体、防灾减灾措施),第三层包括38个指标。这些指标中,大多数为定性指标,最佳的评定办法为专家评价。为了避免人为因素带来的偏差,选用多层模糊综合评价法为评价模型,开发专家评分系统模块支持专家评价过程。将专家评价的分散信息处理为隶属度向量;结合层次分析法确定的指标权重以及加权平均型模糊算子得到各个研究单元对不同土地利用综合灾害风险等级的隶属度;按照最大隶属原则,判定各个评价单元的综合灾害风险等级。(3)上述研究内容中计算模型的构建与GIS技术、DSS技术以及SDSS技术紧密联系。本文分析了SDSS技术在土地利用综合防灾规划中各个环节的应用优势。通过UML统一建模语言及ArcGIS 9.0平台的空间分析数学建模环境进行专业分析模型的构建;采用基于VB语言的ArcObjects开发技术进行空间决策支持模块的修改与定制;利用ArcInfo 9.0为平台,以Oracle 9i为后台数据库;构建了LUCDP-SDSS应用程序。最后对该应用程序的功能、用户接口等进行了介绍。更多还原

【Abstract】 Urban Comprehensive Disaster Prevention Planning(UCDPP) consists of multiple kinds of disasters and urban systems.Based on the present research and practical experience,the components of UCDPP are constructed,in which the Land Use Comprehensive Disaster Prevention Planning (LUCDPP) is the important precondition of the succeeding components of this system.The LUCDPP’s conception,functions and characteristics are discussed,and the research contents are summarized accordingly as:1) combining each disaster’s hazard analysis result and the land attributes to estimate the land’s disaster resistance suitability(e.g.Land’s seismic resistance suitability evaluation),2) combining all the disasters’ hazard analysis results,the land attributes and the attributes of the whole urban system built on the land to research the Land Use Comprehensive Disaster Risk(LUCDR).The research contents have three features:firstly,most data processed are spatial data,secondly,the relevant factors of LUCDR are nonlinear and with randomicity,thirdly the data’s temporal-spatial scope and amount are both great,which limit this problem as a complex unstructured process.With the rapid development of Spatial Technology(ST) and Information Technology(IT),the constructing process for the mathematical model of Spatial Analysis(SA) and Spatial Decision Support System(SDSS) is more professional,modularized,and intelligentized.The following aspects of LUCDPP are researched in details with the combination of LUCDPP theory analysis and SDSS.(1) The seismic disaster is chosen to research the LUCDPP’s first research contents,i.e.the research of Construction Land Seismic Resistance Suitability(CLSRS) evaluation.Based on the present research and practical experience in Urban Seismic Resistance and Hazardous Prevention Planning(USRHPP),seismic disaster prediction and other relevant fields,four indexes(site classification index,liquefaction index,seismic hazard index and ground stability index) are selected to evaluate the CLSRS synthetically.The SA mathematical model of each index and the synthetic evaluating model are built to support the CLSRS evaluation process.●The SA mathematical model for site classification index is based on the geostatistical analyzing process of regionalized variables,Soil Equivalent Shear Wave Velocity (SESWV) and Cover Soil Thickness(CST),which mainly include the fitting model analysis of spatial trend and semivariogram to simulate their spatial distribution raster maps.With the spatial distributive result of SESWV and CST as input,the deterministic and Fuzzy Synthetic Evaluation(FSE) approaches are used separately to calculate the site classification raster map.Practical project data are adopted to illustrate this process.●The probabilistic approach,based on the statistical simulation method(Monte Carlo simulation algorithm),is applied to calculate each bored pile’s liquefaction index.Then the geostatistical analyzing model is used to simulate the spatial distribution of the liquefaction index,i.e.to create each liquefaction grade’s probability raster map.●The Seismic Hazard Analysis(SHA) method FSHASA is constructed on the FSE approach and SA to support the UCDPP and USRHPP,based on the research of present SHA methods.The basic spatial database includes fault,earthquake,geology,bouguer gravity anomaly etc.The models-base includes 11 models for the corresponding 11 indexes of the FSHASA evaluating system and FSE model.The methods-base includes Analytic Hierarchy Process(AHP) for indexes weights,Levenberg-Marquardt for data processing etc.Six districts are selected to test the FSHASA method.●The conception of ground stability is defined first to distinguish from lithosphere stability and foundation stability.The multi-level evaluation indexes system is built, whose first hierarchy indexes include tectonic background,geological disaster,human engineering activities influence,and topographic feature.The FSE approach is used to analyze this indexes system.Based on the results of four indexes for CLSRS,the AHP method is used to synthesize them to get the CLSRS raster map.Considering the limiting factor theory,the raster map is modified.This process is illuminated with practical project data.(2) LUCDR research,i.e.the second research contents of the LUCDPP,is based on each disaster’s hazard analysis result,the land features and various features of the whole urban system built on land to identify the potential synthetic disaster risk.In most cites each disaster’s hazard zonation or other forms of hazard analysis results are available before the UCDPP work.So the emphasis of LUCDR should be the synthetic approach that combines the influencing factors of disaster hazards, land features,various urban system features.The multi-level evaluation indexes system(three levels) is constructed considering every aspect of disaster-inducing factors,disaster-influencing factors and disaster’s bearing-body factors.The first hierarchy includes 5 indexes and the last hierarchy includes 38 indexes.Most of these indexes are qualitative,which are vulnerable by subjective ideas.The multi-level FSE method is adopted to reduce the error caused by subjective factors.The expert appraisal system is built to support the expert appraisal process,and calculate the membership degree value vector.Combined with the weights vector got from AHP method and weighted average fuzzy arithmetic operator,every soil unit’s membership degree value for each LUCDR grade is calculated, and the final LUCDR grade is determined by the principle of maximum membership degree.(3) The GIS,DSS,and SDSS technologies are indispensable for these researches mentioned above.The advantage for the application of SDSS is elaborated.UML and Model Builder(embedded in ArcGIS 9.0) are utilized to build the spatial analysis mathematic model.AO developing technology in VB language is used to develop these models.Finally the LUCDP-SDSS application is built on the ArcInfo 9.0 platform and Oracle 9i RDBMS.Its functions,user interfaces etc.are introduced.更多还原