基于熵值法的洪灾等级评估属性识别模型  

Attribute recognition model for comprehensive evaluation of flood grades based on entropy method

作　　者：杨小玲[1,2] 周建中[1] 丁杰华[3] 张勇传[1] 

机构地区：[1]华中科技大学水电与数字化工程学院,湖北武汉430074 [2]湖北工业大学理学院,湖北武汉430068 [3]中南电力设计院,湖北武汉430071

出　　处：《人民长江》2010年第12期16-19,共4页Yangtze River

基　　金：国家重点基础研究发展计划(973)项目(项目编号:2007CB714107)

摘　　要：鉴于洪灾损失评估中各指标的不相容性和不确定性,将熵值法与属性识别理论建模相结合,建立了基于熵值法的洪灾损失等级评估属性识别理论模型。利用信息熵反映出的数据效用值来刻画指标的权重系数,合理地解决了权重分配问题。该模型以一种新的方法进行洪灾等级划分,能客观地反映洪灾损失的程度。并将该模型应用于1998年特大洪灾实例分析,将评估结果与物元分析法评估结果进行比较,验证了该模型的客观性和合理性。In view of the uncertainty and incompatibility of indexes for flood damage evaluation,we combine the entropy method with the attribute recognition theory,and establish an attribute recognition model for flood damage evaluation based on entropy method.The weight coefficients of this model are characterized by utility value derived from information entropy,which solves the problem of weight allocation properly.The model is based on a new method for flood disaster classification and can reflect flood damage degree objectively.In addition,the calculation result of the flood in 1998 is compared with the results concluded by matter-element analysis method to confirm that the proposed method is objective and reasonable.

关 键 词：洪灾等级 熵 权重 属性识别理论 

分 类 号：P426.616[天文地球—大气科学及气象学]