摘要
文章运用拉曼光谱技术对11种纤维样本进行光谱数据采集并作定性区分,实验发现,在532 nm波长激发下效果最好,且可以用光漂白法消除样本荧光干扰;对采集的纤维光谱曲线作定性分析,统计得到拉曼特征峰位表;对光谱数据作主成分分析发现,降至15维时已有累计99.34%的信息贡献率。降至三维的可视化分析表明不同纤维拉曼光谱具有较高的区分度,同种纤维光谱数据有聚类特性。
This paper uses Raman spectroscopy to collect spectral data from 11 types of fibers and make qualitative distinctions.The best excitation light source is found at 532 nm,and at this wavelength,the fluorescence interference can be removed by photobleaching;the collected fiber spectral curve is qualitatively analyzed,and the Raman characteristic peak position table is obtained statistically;principal component analysis is performed on the spectral data,it is found that when reduced to fifteen dimensions,still have cumulative 99.34%information variance.Reduced to 3D visual analysis,the Raman spectra of different fibers have high differentiation,and the spectral data of the same fiber have clustering characteristics.
作者
张辉
汤泽辉
黄毅涵
Zhang Hui;Tang Zehui;Huang Yihan(Zhejiang Police College,Hangzhou 310053,China;Qiantang Branch of Hangzhou Police Bureau,Hangzhou 310053,China)
关键词
拉曼光谱
纤维
主成分分析
特征峰位
Raman spectrum
fiber
principal component analysis
characteristic peak position