基于样本加权模糊聚类的土壤重金属污染溯源  

Soil heavy metal pollution traceability based on sample weighted fuzzy clustering

作　　者：陈志奎[1,2] 李丽方 林聃 CHEN Zhikui;LI Lifang;LIN Dan(College of Software Technology,Dalian University of Technology,Dalian Liaoning 116620,China)

机构地区：[1]大连理工大学软件学院,辽宁大连116620 [2]大连理工大学辽宁省泛在网络与服务软件重点实验室,辽宁大连116620

出　　处：《工业安全与环保》2023年第6期95-99,共5页Industrial Safety and Environmental Protection

基　　金：国家自然科学基金(62076047);大连创新基金(2021JJ12SN44);中央高校重点实验室基金(DUT20LAB136)。

摘　　要：由于土壤数据的复杂性,现有的方法难以对土壤重金属进行高效溯源,为了解决这个问题,基于样本加权的模糊聚类算法建立了一种新的重金属污染溯源模型,该模型使用一种新的相似度计算方法来构建模糊相似矩阵,将聚类结果与土壤重金属污染来源表相结合,可以快速确定土壤重金属的污染来源。通过与欧式距离、皮尔逊相关系数、余弦相似度等相似度计算方法在天津市武清区数据集上的对比实验验证了所提出方法的有效性与可用性,结果表明,该方法可用于土壤重金属污染源解析方面的工作。Because of the complexity of the soil data,however,the existing methods of soil heavy metal hard effectiv ely,in order to solve this problem,a new heavy metal pollution traceability model based on sample weighted fuzzy clustering algorithm is established.The model uses a new similarity calculation method to construct the fuzzy similar matrix,the clustering results combined with soil heavy metal pollution source table,the pollution sources of heavy metals in soil can be determined quickly.The validity and availability of the proposed method were verified by com paring the similarity calculation methods with Euclidean distance,pearson correlation coefficient and cosine similarity on the data set of Wuqing District,Tianjin.The results show that the proposed method can be applied in the analysis of soil heavy metal pollution sources.

关 键 词：重金属污染溯源 模糊聚类 样本加权 数据挖掘 

分 类 号：X53[环境科学与工程—环境工程]