【作者】 陈静；

【导师】 胡振琪；

【作者基本信息】 中国矿业大学（北京） ， 地图制图学与地理信息工程， 2009， 博士

【摘要】 运用地理信息系统、多准则决策理论、模糊数学等理论知识,研究了空间多准则评价方法在城市防灾决策分析,尤其是城市土地防灾适宜度评价方面的技术方法。分析了OWA方法与布尔决策和权重线性叠加(WLC)等其它基于GIS的常用多准则评价方法在决策策略上的区别。OWA方法的优势是在决策过程中考虑了不同的空间风险态度,更为符合实际决策境况。应用的关键是次序权重和准则权重的计算,通过应用修正模糊量化方法求次序权重,应用层次分析程序(AHP)构建比较矩阵来计算准则权重。分析比较了决策风险与指标补偿程度关系。准则地图是决策分析的基础,需要专业模型和GIS紧密结合,探讨了应用BP人工神经网络模型开发准则图层的方法。以唐山市为例,基于GIS-OWA方法建立城市土地防灾适宜度评价的决策分析模型,计算了唐山市中心区城市建设用地防灾适宜度。运用D-S证据理论探讨间接证据引起的决策不确定性问题。最后编制基于综合防灾的唐山市土地开发利用管理信息系统,为决策者提供决策支持。更多还原

【Abstract】 Based on the theories such as GIS, decision-making technology and fuzzy mathematics, investigation and research are done on the spatial multi-criteria evaluation methods applicable to problems of urban disaster prevention decision, especially to evaluate land-use suitability affected by geological disaster factors. The difference in decision strategy between OWA and other GIS based MCE methods, such as Boolean overlay and weighted linear combination (WLC) which are often used in decision-making, is analyzed. The main advantage of the method proposed is the incorporation of spatially variable attitude to risk into the decision-making process, which Simulate decision situation in fact. How to calculate order weights and criteria weights are the key of problem, modified fuzzy-quantified methods to get order weights and analytical hierarchy program (AHP) to construct comparison matrix for criteria weights are introduced. Criteria maps of spatial information are base of decision analysis, and can be got by special model combining with GIS. The paper use artificial neural network methods to forecast Karst collapse hazard, then take Tangshan as an example to analyze its application in urban land-use suitability affected by multi-geological disaster factors, which are investigated after building land-use information system in Tangshan city, getting spatial data and constructing decision-making model on land-use suitability of disaster prevention. The paper discusses a D-S evidence approach to decision-making uncertainty roused by indirect evidence.更多还原