区域环境质量评价与模糊聚类分析

近年来区域环境质量评价虽进展迅速,但加权与分级仍为其中最无把握、且最易引起争议的问题,迄今仍未妥善解决。本文基于上海某区的环境质量评价提出了一种新方法的应用,即模糊聚类分析,它自然而简捷地解决了上述问题。首先,它在被分级的元素之间建立一个模糊相似关系,再通过模糊矩阵的复合运算,由细到粗地逐步绘出一棵动态聚类树,最后根据地方环境状况确定一个λ值以后,就可获得一个客观的模糊聚类的质量分级。通过将其结果与系统分级的比较。显示了两者的某些差异。

REGIONAL ENVIRONMENTAL QUALITY ASSESSMENT AND FUZZY CLUSTER ANALYSIS

In spite of rapid progress achieved recently on research of regional environmental quality assessment, problems of weightedness and classification have never been solved so far. This paper presents a new approach-fuzzy cluster analysis, based upon environmental quality assessment work done in a certain district of Shanghai, which deals with the above problems in an easier and more natural way. First, there exists a similarity relation of fuzzy among the elements being classified, and then a dynamic clustering tree from thin to thick can be gradually drawn up. Finally, after a proper threshold determined, i. e. value A according to the local condition, an objective classification of fuzzy cluster can be obtained. The result is used to be compared with those by systematic classification and there shows some difference.