摘要
综述化学计量学结合拉曼光谱在生物材料检测中的应用。报道利用基线校正方法(导数、曲线拟合、小波变换)、归一化方法(矢量归一化、峰高归一化)、复合预处理等方法,可以提高信噪比,恢复失真的信号,实现谱图峰位的正确识别;基于主成分分析、偏最小二乘、聚类分析、目标因子分析、人工神经网络、支持向量机等算法构建化学计量模型,利用模型对拉曼光谱定性或定量分析,分别得到可信的结果;提出未来化学计量建模的方向。
The paper reviews the application of the combination of chemometrics with Raman spectroscopy to detect biomaterials. The preprocessing methods reported in the paper such as baseline correction methods （derivatives,polynomial fitting,wavelet transform）, normalization methods （vector normalization, peak normalization）, combination methods, can enhance signal-to-noise, improve the distorted signal and achieve the correct identification of spectra peaks. Reliable results can be obtained using the chemometrical model based on the algorithms of principal component analysis,partial least squares,cluster analysis, target factor analysis, artificial neural network and support vector machines in qualitative or quantitative analysis of Raman spectra. The paper also proposed the direction of ehemometrieal models development in the future.
出处
《光谱实验室》
CAS
CSCD
北大核心
2010年第6期2442-2448,共7页
Chinese Journal of Spectroscopy Laboratory
基金
广西自然科学基金(桂科自0991078)
广西自然科学基金(桂科回0731013)
广西师大校内项目
广西科学院基本科研业务费资助项目(08YJ16WL02)
关键词
化学计量学
拉曼光谱
光谱预处理
化学计量建模
Chemometrics
Raman Spectroscopy
Spectral Preprocessing
Chemometrical Modelling
