灰色聚类模型的改进及应用研究

Improvement and Application of Grey Clustering Model in Atmospheric Quality Comprehensive Evaluation

对系统进行质量等级综合评价,应充分考虑各影响因子之间的相互作用关系.基于指数型白化函数的灰色聚类改进模型避免了经典聚类模型存在零权重现象的缺陷,使得指标样本值与各级非零权重之间建立了一一对应关系,并将该模型用于计算各单项因子的最大控制值或最小削减量.实例分析了2004年杭州市环境空气质量综合等级为Ⅲ级,并计算了综合等级要达到Ⅱ级标准时,单项指标SO2、NO2和PM10的浓度最小削减率分别为22.69%、27.9%和21.58%.

A new methodology is presented for comprehensive evaluation of system quality grade based on the exponential whiteness function. Zero weight problems, which occur with the classical grey clustering model, can be solved. A one to one corresponding relationship between value of the index sample and nonzero weight of various grades is established in the new model. As an application, this model can conclude the formula of each single index to work out the maximal value to control or the minimal quantity to cut down. The comprehensive evaluation result of Hangzhou＇s atmospheric environmental quality in 2004, indicates the grade is Ⅲ. With the aim at the grade Ⅱ ,it also works out the minimal rate to cut down of SO2,NO2 and PM10 are 22.69%, 27. 9% and 21.58% respectively.