哈尔滨大气中多环芳烃污染的FCM聚类算法

摘要

To improve the analysis ability of residue levels of persistent toxic substances（PTS）in environ-ment,this paper investigated the distribution characteristics of air concentrations of polycyclic aromatic hydro-carbons（PAHs）during the Spring of 2007（January to April）in 4 different functional areas（urban,subur-ban,rural,and remote areas）in and near the Harbin metropolitan by using the fuzzy C-means clustering al-gorithm,and got the cluster centers of different samples and a membership matrix which offered quantity foun-dation for analysis of samples description.Results showed that the contamination level in the urban industrial area was the highest,followed by those in the other urban places and rural area,and the contamination levels in the remote area and in the places on the windward were the lowest.PAHs in Harbin urban atmospheric were mainly from coal-burning and industrial emissions,those in rural areas were mainly from the burning of crop stalks.PAHs in air in this region have imposed a potential threat on human health.%为提高环境中持久性有毒物质残留水平的分析能力,以2007年1～4月哈尔滨市4个不同功能区域（市区—郊区—农村—偏远地区）的大气中多环芳烃（PAHs）的监测质量浓度为数据样本,运用模糊C均值聚类算法作为样本聚类的方法,研究该地区2007年春季大气中PAHs的分布特征,并得到不同样本的聚类中心以及样本的隶属度矩阵,为样本的特征分析提供量化依据.分类结果发现：城市工业区的污染状况最为严重,农村地区和介于市区和工业区的居民区的区域次之,而远离污染源的市内居民区、城市上风向和偏远地区污染较轻.分析结果表明,哈尔滨市区大气中PAHs主要来自于燃煤和工业排放,农村大气中的PAHs主要来自于农作物秸秆的燃烧.市区和农村地区大气中PAHs对于人类的健康存在较大潜在威胁.