摘要
采用近红外漫反射光谱法测得不同种类纺织纤维的近红外光谱图,运用Bruker Opus定性分析软件对相应的近红外谱图进行化学计量法处理并建立鉴别模型,测试了模型的准确性。结果表明:将获取的近红外谱图组合为一个主库和两个子库,数据预处理方法为一阶导数和预矢量归一化,主库谱图的运算方法为标准算法,两个子库的运算方法为因子算法,可以得到满意的纺织纤维鉴别模型。
Near infrared spectra of different textile fibers were gained by near infrared diffuse reflectance spectroscopy.Bruker OPUS qualitative analysis software was applied to carry out stoichiometry treatment of infrared spectra and establish the identification model.Besides,the model,odel accuracy was tested.The results show that the satisfying textile fiber identification model can be gained under the following conditions:near infrared spectra of different textile fibers are combined as a main library and two sub-libraries;data preprocessing methods include first derivative and pre-vector normalization;the operation method of main library is canonical algorithm;the operation method of two sub-libraries is factor algorithm.
出处
《现代纺织技术》
2018年第2期62-65,共4页
Advanced Textile Technology
关键词
近红外漫反射光谱法
纺织纤维
模型
鉴别
near infrared diffuse reflectance spectroscopy
textile fibers
model
identification
