基于近红外光谱法快速检测毛涤织物中的纤维含量

Fast Detection of Fiber Content in Wool and Polyester Fabric by Near-Infrared Spectroscopy

为提高纺织面料成分的检测水平和分析速度,以134个毛涤混纺织物为研究对象,自行设计采样装置,采集其近红外光谱;然后,经一阶导数等光谱预处理方法,结合偏最小二乘法,建立了毛涤混纺织物中毛含量近红外的定量校正模型。为验证模型的实用性,对129个毛涤样品进行了预测,将预测结果进行方差分析,得到该方法与标准SN/T 0464-2003方法的结果不存在显著性差异的结论,并据此制定了纺织品原料组分近红外快速检测初筛方法的标准操作规程,经实际应用证明效果显著。

To speed up the analysis and testing level of textile fibers, 134 samples of wool and polyester blended fabrics were taken as research objects for analysis in this paper. The near-infrared（NIR） spectra of samples were collected using self designed accessory, and then first order derivative were used to pretreat the spectra. Quantitative Correction Model for wool content were established with Partial Least Square（PLS）. To verify the models, wool content prediction of 129 samples was conducted. The variance analysis shows that there is no significant difference between NIR method and national standard method SN/T 0464-- 2003. The standard operation procedure（sop） worked out on the above result is a rapid near-infrared detec- tion method for textile material components, which is proved an effective method.