• 关键词：模糊聚类  FCM  水质评价  海河

• 基金项目：国家自然科学基金资助项目(No50778009)

• 邹志红
• 北京航空航天大学经济管理学院, 北京 100083

• 云逸
• 北京航空航天大学经济管理学院, 北京 100083

• 王惠文
• 北京航空航天大学经济管理学院, 北京 100083

• 摘要：将两阶段模糊法用于海河水系监测断面的水质评价.与传统模糊聚类法不同,该方法首先将国家地表水指标转化为等级打分,再构造模糊隶属函数,确定评价对象对标准水质水平的隶属度;之后运用改进的模糊c-均值法(FCM法)选取适当模糊指标和先验类中心,对评价对象进行聚类,得到替代原观测值的一个全局水质指标,数值在0～100之间,以利于直观评价.计算了海河水系2004～2006年4个断面的432个数据,考察了季节波动规律和不同年度变化的情况,并对水质进行了综合评价.两阶段模糊法对水质评价综合指标有较好的分类能力,能够得到十分科学的评价结果,且计算过程相对FCM法简便,可以运用于河流水质综合评价.

• Abstract：Two-stage fuzzy set theory was used to evaluate the water quality in Haihe River system monitoring sections. Unlike traditional fuzzy clustering methods, the national surface water indexes were first transferred into ranking marks. Fuzzy membership functions were formed to determine the membership degree of evaluation objects for each standard water level. The improved fuzzy c-means (FCM) methodology was used afterwards, and the proper fuzzy index and transcendent class center were chosen for classification. Then an overall index, the River quality index (RQI), was constructed to represent the original monitoring data. The value of RQIwas designated by the interval [0, 100], to make intuitional evaluation. Altogether, 432 data points from 4 monitoring sections in the Haihe River system during the years 2004～2006 were calculated. Seasonal fluctuations and different situations among years were considered, and the water quality was synthetically evaluated. Two-stage fuzzy set theory is capable of classifying water quality. Accurate scientific results are obtained, and the calculation process is relatively simple compared with usual TCM, so it can be applied in water quality evaluation in Chinese river systems.

• Key words：fuzzy clustering  FCM  water quality evaluation  Heihe

• 
  摘要点击次数： 2281 全文下载次数： 2972