主成分分析法用于土壤样品分类  

CLASSIFICATION OF THE SOIL SAMPLES BY APPLICATION OF PRINCIPAL COMPONENT ANALYSIS

作　　者：李芳清[1] 倪永年[1] 杨水平[2] 

机构地区：[1]南昌大学化学系,江西南昌330047 [2]东华理工学院化学系,江西抚州344000

出　　处：《南昌大学学报（理科版）》2004年第2期140-143,共4页Journal of Nanchang University(Natural Science)

基　　金：国家自然科学基金资助项目(20065002);江西省自然科学基金资助项目(0320014);江西省教育厅科研基金资助项目(03-12)

摘　　要：采用火焰原子吸收光谱法对江西省南丰县南丰蜜桔种植区和非种植区土壤中锰、铜、锌、钴、镍5种元素的含量进行测定;用电位滴定法测定上述土壤样品的酸度;用灼烧法测定其中有机物质的含量;用干燥法测定自然水的含量。进而采用主成分分析法(PCA)对所测得的土壤样品的8个变量进行分类研究,发现蜜桔种植区和非种植区的样品能得到较好的分离识别,证明南丰蜜桔的质量与土壤中的微量元素的含量、有机物质的含量以及土壤的酸度之间存在相关性。The classification of soil samples obtained from orange farm Nanfeng countryside in Jiangxi Province,as well as samples obtained from other area (total 20 samples) has been studied. Five metal elements (copper, zinc, cobalt, nickel and manganese) were determined with flame atomic absorption spectrometry (FAAS), the acidity of soil was determined by direct potential method, the moisture of the soil was determined by the drying method, and the organic contents was determined by burn method. The pattern recognition based on the principal component analysis (PCA) was applied to the measured data of the soil samples. The result indicated that abundance statistics characters can be obtained and the samples obtained from different area can also be distinguished effectively when the variables were selected appropriately.

关 键 词：主成分分析 土壤 分类 

分 类 号：O657.32[理学—分析化学]