佳木斯大学学报(自然科学版)
佳木斯大學學報(自然科學版)
가목사대학학보(자연과학판)
JOURNAL OF JIAMUSI UNIVERSITY (NATURAL SCIENCE EDITION)
2015年
2期
286-289
,共4页
刘高杰%李云霞%张二喜%白林
劉高傑%李雲霞%張二喜%白林
류고걸%리운하%장이희%백림
重金属%反距离加权插值%空间分布%地累积指数法%基于神经网络的K-means聚类分析
重金屬%反距離加權插值%空間分佈%地纍積指數法%基于神經網絡的K-means聚類分析
중금속%반거리가권삽치%공간분포%지루적지수법%기우신경망락적K-means취류분석
heavy metal%inverse distance weighted interpolation method%spatial distribution%geoaccum-clation index%K -means clustering analysis based on neural network
根据湖南省多目标地球化学调查资料,研究了易家湾地区表层土壤中重金属元素As, Cd,Hg,Pb,Zn 的污染状况并寻找污染源。利用地累积指数法对各元素污染程度做出评价,运用反距离加权插值法,对元素的空间分布进行分析,了解元素在各采样点属性值,最后由基于神经网络的K-means聚类分析确定元素的污染源。结果表明Cd在采样区存在强度以上的污染,Hg存在中-强度污染的采样点达20%,而As,Pb和Zn均存在中度以下污染,并且基于神经网络的K-means聚类分析不仅达到去噪的目的,在易家湾污染源分析中是合理的。
根據湖南省多目標地毬化學調查資料,研究瞭易傢灣地區錶層土壤中重金屬元素As, Cd,Hg,Pb,Zn 的汙染狀況併尋找汙染源。利用地纍積指數法對各元素汙染程度做齣評價,運用反距離加權插值法,對元素的空間分佈進行分析,瞭解元素在各採樣點屬性值,最後由基于神經網絡的K-means聚類分析確定元素的汙染源。結果錶明Cd在採樣區存在彊度以上的汙染,Hg存在中-彊度汙染的採樣點達20%,而As,Pb和Zn均存在中度以下汙染,併且基于神經網絡的K-means聚類分析不僅達到去譟的目的,在易傢灣汙染源分析中是閤理的。
근거호남성다목표지구화학조사자료,연구료역가만지구표층토양중중금속원소As, Cd,Hg,Pb,Zn 적오염상황병심조오염원。이용지루적지수법대각원소오염정도주출평개,운용반거리가권삽치법,대원소적공간분포진행분석,료해원소재각채양점속성치,최후유기우신경망락적K-means취류분석학정원소적오염원。결과표명Cd재채양구존재강도이상적오염,Hg존재중-강도오염적채양점체20%,이As,Pb화Zn균존재중도이하오염,병차기우신경망락적K-means취류분석불부체도거조적목적,재역가만오염원분석중시합리적。
According to the data obtained from the multipurpose regional geochemical survey of Hunan province, the pollution condition and source of surface soil heavy metals , such as As, Cd, Hg, Pb and Zn in Yi-jiawan area were analyzed .Pollution level of heavy metals was accumulated by geoaccumulation index , inverse distance weighted interpolation method was used to make evaluation on elements and acquaintance values of sam -pling points , and Pollution source can be determined by K -means clustering analysis based on neural network . Conclusion can be got as this: the pollution degree of Cd outnumber intensity in the sampling area , whic ac-counts for 20%of the sampling points and Hg exits middle -intensive pollution , while the As , Pb and Zn all ex-ist pollution under the moderate pollution , and K-means clustering analysis based on neural network can not on-ly achieve the goal of denoising , but is reasonable for the pollution source analysis in Yijia wan area .