偏最小二乘回归
主成分分析
主成分回归
欧几里德距离
水质
吸收(声学)
回归分析
统计
分析化学(期刊)
化学
数学
生物系统
计算机科学
色谱法
人工智能
材料科学
生物
复合材料
生态学
作者
Yanpu Zhao,Xia Li,Xiao Liu,Pengfei Dong,Lingli Wang,Xianquan Wang
出处
期刊:PubMed
日期:2016-11-01
卷期号:36 (11): 3592-6
被引量:4
摘要
Using the UV absorption spectrum to detect Organic pollutants content in water has become one of the most important methods for real-time online monitoring in the field of water quality inspection, however, the water complex and unstable components often bring much uncertain offset to the standard test. In this paper, water samples were classified firstly by analyzing UV absorption spectrum ranging from 200 nm to 400 μm including the organic substances, through the way of combining principal component analysis (PCA) with Euclidean distance. In this paper, we compared the Principal component analysis combined with partial least squares regression (PCA-PLSR) and the direct multi-wavelength absorption models combined with partial least squares regression (MWA-PLSR), not only for the real water sample but also for the analysis of different concentrations of COD standard solution. The result indicates that the measurement errors of the PCA is less than 5%, it is the smallest by using the first and second principal components as regression parameters for PLSR. Using the methods above can simultaneously achieve to classify of water samples and to measure the concentration of water quality parameters more accurately.
科研通智能强力驱动
Strongly Powered by AbleSci AI