【作者】 付博；

【导师】 姜琦刚；

【作者基本信息】 吉林大学 ， 地学信息工程， 2011， 博士

【摘要】 城市的宜居性是指城市在经济、社会、文化、环境方面协调发展,人居环境良好,能够满足居民物质和精神生活需求,适宜人类生活、工作和居住的性质。长春作为吉林省的省会及吉林省政治、经济、文化和交通中心,自20世纪80年代起进入快速发展期,在城市经济建设方面取得巨大成就的同时,出现了城住房紧张、交通堵塞、环境污染等一系列城市问题。在这一社会背景下,建设宜居城市显得尤为必要与紧迫。本论文进行长春市城区的城市宜居性评价研究,以GIS与遥感作为技术手段,提取制约长春市宜居性的因素因子,建立评价指标模型,采用均方差——主成分分析法,确定指标因子权重,运用多目标线性加权函数——综合比较法进行长春市宜居性评价,并与问卷调查获得的居民主观满意度进行了耦合分析。论文具体包括如下研究内容：1、论文根据长春市城市规模和具体条件,针对多来源、多形式、多尺度的基础数据源,利用GIS和遥感方法,进行了综合集成与融合处理,并采用栅格化方法、空间插值法、窗口分析等方法,对长春市宜居性主要影响因子进行了提取并建立了因子空间数据库。2、论文总结了国内外宜居性评价理论和方法,并以此为基础,建立了由目标层、准则层、因素层、指标层构成的宜居性评价指标体系,采用均方差——主成分分析法确定各层指标权重,避免了定权过程中人为因素的影响,同时按多目标线性加权函数法,在获取各层次量化指标值的同时,进行了研究区自然环境条件、人文环境条件及研究区总体宜居性的评价分析。3、论文在宜居性评价分析中,采用栅格单元作为建模分析的基本单元,兼顾栅格数据量和数据精度,通过数学分析方法确定了栅格单元的最适宜大小为100m×100m。4、论文通过数学方法,将宜居性居民满意度调查问卷中获取的定性指标值进行了量化处理,并通过主成分分析法确定了各指标权重以计算便利性、舒适性、健康性和安全性指标的主观满意度,并与评价指标体系中的人文环境对应指标进行多元回归分析以获取人文环境各指标对宜居性居民主观满意度的影响强弱及量级,确定了宜居性建设中高度关注、优先改进、维持优势及无关紧要的各方面。5、论文根据各栅格单元的总体宜居度,将研究区30552个栅格划分为Ⅰ类、Ⅱ类、Ⅲ类、Ⅳ类宜居区,同时根据各街道总体宜居性的主观满意度分值将研究区46个街道划分为一级、二级、三级、四级区域,通过进行以栅格为单元的长春市宜居性和以街道为单元的宜居性居民主观评价的耦合分析,总结了各级居民满意度区域内宜居性环境条件的差异。本论文利用遥感和GIS技术,为城市宜居性评价提供了多尺度的数据源、空间分析及可视化表达方法,论文的特色和创新在于：1.在GIS与遥感技术平台支持下,实现了多源数据的综合集成与融合处理,并通过栅格化、空间插值、窗口分析等方法构建了基于像元的城市宜居性评价指标空间数据库,较好地解决了社会统计数据无法与空间位置准确匹配及难以实时更新的问题,与以往研究相比较,能够实现较高分辨率的宜居度空间格局及区域差异的可视化表达。2.运用均方差—主成分分析法确定评价指标体系中指标因子的权数,建立了多要素、综合定量的长春市宜居性评价模型。利用主成分回归分析法、多目标线性加权函数、耦合分析等方法,将社会调查数据与城市环境背景数据结合,为城市宜居性建设提供量化的参考。更多还原

【Abstract】 Urban livability is the balanced developments of a city in economy, society, culture, environment and a good inhabitable environment meets the needs of the residents in material and spiritual life and is suitable for human living, working and living. Changchun, the capital and the political, economic, cultural and transportation center of Jilin Province, has developed rapidly since 1980s. However, the city hustle, traffic jams, environmental pollution and a series of urban problems have occurred with the great achievements in economic growth. Under this social environment, construction of urban livability is particularly necessary and urgent. The evaluation of urban livability in this dissertation is conducted by taking GIS and Remote Sensing as technical means to extract the factors of urban livability and establish the index model of the evaluation, the combination of Mean Variance and Principle Component Analysis as the method to determine the index weights of the factors, and the combination of Multi-objective Linear Weighting Function and Comprehensive Comparison to evaluate the livable degree. The coupling analysis with subjective satisfaction of the inhabitants from questionnaire is also conducted in the dissertation.The contents of the research included in this dissertation are as follows:1. With the help of GIS and Remote Sensing techniques, the multi-source, multi-form and multi-scale basic data sources are integrated and fused synthetically and the major factors of urban livability are extracted by Raster Data Method, Spatial Interpolation, Window Analysis to establish the spatial database of the factors.2. Based on the summarization of evaluation theories and methods from home and abroad, the index system of livability evaluation consisting of the target layer, rule layer, factor layer, indicator layer is established and the index weights are determined by the combination of Mean Variance and Principle Component Analysis to avoid the influence of human factors. By the combination of Multi-objective Linear Weighting Function and Comprehensive Comparison, the quantitative values of the indicators are obtained, and the evaluation analysises of the natural environmental conditions, humanistic environmental conditions and the overall livability of the study area are conducted.3. In the evaluation and analysis of the urban livability, grid cells are taken as the basic units of modeling and analysis. Taking into account the amount and the accuracy of the raster data, the optimum grid cell size of 100m×100m is determined by mathematical analysis.4. Qualitative index values obtained by the resident satisfaction questionnaire of the urban livability are quantified by mathematical method and the index weights are determined by PC A to evaluate the subjective satisfaction scores of convenience, amenity, health and safety indicators. The multiple regression analysis with the corresponding indicators of the humanistic environment is conducted to obtain the strong or weak influence and its magnitude. The aspects of high attention, priority improvement, domination maintenance and no consequence are also fixed on.5. The 30552 grids of the study area are divided intoⅠ,Ⅱ,Ⅲ,Ⅳgrade of the livable regions according to the overall livable degrees of the each grid cell and the 46 streets are divided into the primary, secondary, third, fourth-level areas according to livable satisfaction scores of the streets. The impacts of the objective livability conditions on the resident satisfaction are summarized by the coupling analysis between the objective livable degrees taking grid cells as units and the subjective livability evaluation taking the streets as units.By using GIS and Remote Sensing technique, multi-scale data sources, spatial analysis and visualization methods are provided to the evaluation of urban livability. The peculiarity and innovation of this dissertation lie in:1. The multi-source data are integrated and fused synthetically with the help of GIS and Remote Sensing techniques and the spatial indicator database of livability evaluation based on the pixels is established by the Raster Data Method, Spatial Interpolation and Window Analysis. The issues of the social statistics that are not able to match the spatial location accurately and difficult to update real-timely can be better resolved. Compared with the previous studies, the spatial patterns of the livability of the higher resolution and visual expression of regional differences can be realized in this dissertation. 2. The index weights of the indicator system are determined by the combination of Mean Variance and Principle Component Analysis and the evaluation model of multi-elements and comprehensive quantization is established. Methods of the Principal Component Regression Analysis, Multiple Objective Linear Weighting Functions, Coupling Analysis are used to provide the quantitative references of urban livability construction by the integration of social investigation data and the background data of the urban environment.更多还原