展开基于独立成分分析(ICA)对复杂有机物混合体系盲源解析的系统研究。通过确立合理的独立成分数目的方法,分别利用模型分离信号的重构信号与原始信号之间的均方根误差、主成分的方差贡献率对独立成分数目的选择进行优化。综合3种ICA算法完成了以下研究:(1)含硝基苯等多种环境有机污染物混合质谱信号的源解析,其中Kernel-ICA提取的独立成分与实际的源信号之间具有较高的相关性,R平均值(标准差)为0.8697(0.10),可以满足定性识别的要求;(2)复方氨酚烷胺药物的紫外光谱信号中特征组分信息的提取,Kernel-ICA对药物主要成分对乙酰氨基酚提取的有效度最大。该研究工作为构建有机物体系最优盲源解析模型提供了理论支持,为实际环境样本中有机污染物的源解析、药物有效成分的提取提供了有效的手段。
Abstract
A systematic study on blind source analysis of complex organic mixture system based on independent component analysis (ICA) was carried out. By establishing a reasonable number of independent components, the selection of the number of independent components was optimized by the root mean square error between the reconstructed signal of the separation signal and the original signal, and variance contribution rate of principal components.The following studies have been completed by integrating three ICA algorithms: (1) The source analysis of the mixed mass spectrum signals of various environmental organic pollutants including nitrobenzene, among which the independent components extracted by Kernel ICA have a high correlation with the actual source signals, and the average value (standard deviation) of the substance R is 0.8697 (0.10), which can meet the requirements of qualitative identification; (2) The extraction of characteristic component information in the ultraviolet spectrum signal of Compound Paracetamol and Amantadine Hydrochloride, Kernel ICA is the most effective algorithm for extracting the main components of drugs. The aboved research work provides theoretical support for the construction of the optimal blind source analysis model of organic matter system, and provides an effective means for the source analysis of organic pollutants in actual environmental samples and the extraction of effective components of drugs.
关键词
计量学;独立成分分析;有机污染物 /
独立成分数目;盲源信号分离;信息提取
Key words
metrology /
independent component analysis /
organic pollutants /
number of independent components /
blind source signal separation /
information extraction
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
［1］张立国, 张玉曼, 金梅, 等. 基于盲源分离的运动想象脑电信号特征提取方法的研究［J］. 计量学报, 2015, 36(5): 535-539.
Zhang L G, Zhang Y M, Jin M, et al. A Research on the Method of Motor Imagery EEG Feature Extraction Based on Blind Source Separation［J］. Acta Metrologica Sinica, 2015, 36(5): 535-539.
［2］樊凤杰,白洋,纪会芳.基于EEMD-ICA的脑电去噪算法研究［J］. 计量学报, 2021, 42(3): 395-400.
Fan F J, Bai Y, Ji H F. Denoising Method of EEG Signal Based on EEMD-ICA［J］. Acta Metrologica Sinica, 2021, 42(3): 395-400.
［3］邢婷婷,关阳,孙登云,等.基于FSDPC_Otsu算法的滚动轴承故障研究［J］. 计量学报, 2021, 42(11): 1466-1471.
Xing T T,Guan Y,Sun D Y,et al. Rolling Bearing Fault Diagnosis Based on Clustering by Fast Search and Find of Density Peaks Combined Otsu Method［J］. Acta Metrologica Sinica, 2021, 42(11): 1466-1471.
［4］Bell A J, Sejonowski T J. An information-maximization approach to blind separation and blind deconvolution［J］. Neural Computation, 1994, 7(6): 1129-1159.
［5］Hyvrinen A, Karhunen J, Oja E. Independent Component Analysis［M］. Hoboken:John Wiley & Sons.2001: 23-27.
［6］Ai-Qazzaz N K, Alyasseri Z A A, Abdulkareem K H, et al. EEG feature fusion for motor imagery: A new robust framework towards stroke patients rehabilitation［J］. Computers in Biology and Medicine, 2021, 137: 104799.
［7］Loudiyi M, Karoui R, Rutledge D N, et al. Fluorescence spectroscopy coupled with independent components analysis to monitor molecular changes during heating and cooling of Cantal-type cheeses with different NaCl and KCl contents［J］. Science of Food and Agriculture, 2018, 26(33): 61-62.
［8］Monakhova Y B, Mushtakova S P. Multicomponent quantitative spectroscopic analysis without reference substances based on ICA modelling［J］. Analytical and Bioanalytical Chemistry, 2017, 409(13): 3319-3327.
［9］张鹏伟. 采用ICA的公共信道多干扰源信号的自动识别方法［J］. 智能系统学报, 2013, 8(3): 277-282.
Zhang P W. Automatic recognition of multi-interference source signals in the common channel based on independent component analysis［J］. CAAI Transactions on Intelligent Systems, 2013, 8(3): 277-282.
［10］Monakhova Y B, Rutledge D N. Independent components analysis (ICA) at the “cocktail-party” in analytical chemistry［J］. Talanta, 2019, 208(1): 120451.
［11］石玉虎, 曾卫明, 邓金, 等. 基于多目标优化约束独立成分分析方法的fMRI数据分析研究［J］. 中国生物医学工程学报, 2021, 40(1): 19-32.
Shi Y H, Zeng W M, Deng J, et al. Study on fMRI Data Analysis Based on Multi-Objective Optimization CICA［J］. Chinese Journal of Biomedical Engineering, 2021, 40(1): 19-32.
［12］田立勇, 张一辙. 基于局部均值分解与快速独立成分分析的潜水泵故障诊断［J］. 计量学报, 2020, 41(5): 585-591.
Tian L Y, Zhang Y Z. Fault Diagnosis of Submersible Pumps Based on Local Mean Decomposition and Fast Independent Component Analysis［J］. Acta Metrologica Sinica, 2020, 41(5): 585-591.
［13］Bach F R, Jordan M I. Kernel Independent Component Analysis［J］. Journal of Machine Learning Research, 2003, 3(1): 1-48.
［14］Wang G Q, Cai W S, Shao X G. A primary study on resolution of overlapping GC-MS signal using mean-field approach independent component analysis［J］. Chemometrics & Intelligent Laboratory Systems, 2006, 82(1-2):137-144.
［15］谷凤伟, 陆军, 夏桂华. facenet皮尔森判别网络的人脸识别方法［J］. 智能系统学报, 2022, 17(1): 107-115.
Gu F W,Lu J, Xia G H. Face recognition method based on facenet Pearson discrimination network［J］. CAAI Transactions on Intelligent Systems, 2022, 17(1): 107-115.
［16］庞佳烽, 汤谌, 李艳坤, 等. 中红外光谱联合模式识别鉴别奶粉中三聚氰胺［J］. 光谱学与光谱分析, 2020, 40(10): 3235-3240.
Pang J F, Tang C, Li Y K, et al. Identification of Melamine in Milk Powder by Mid-InfraredSpectroscopy Combined With Pattern Recognition Method［J］. Spectroscopy and Spectral Analysis, 2020, 40(10): 3235-3240.
［17］Li Y K, Zeng X C. Serum SELDI-TOF MS analysis model applied to benign and malignant ovarian tumor identification［J］. Analytical Methods, 2016, 8(1): 183-188.
［18］Wang G Q, Sun Y A, Ding Q Z, et al. Estimation of source spectra profiles and simultaneous determination of polycomponent in mixtures from ultraviolet spectra data using kernel independent component analysis and support vector regression［J］. Analytica Chimica Acta, 2007, 594(1): 101-106.
基金
中央高校基本科研业务费(2017MS135);国创计划资助项目(X2021-284)
PDF(99024 KB)
