毕设论文终稿.doc

纺织品中棉含量的近红外光谱快速测定摘要本文建立了分析精度较高的用于检测纺织品里含棉量的近红外光谱分析模型。选择了 138 个纺织品样品作为对象,利用布鲁克 TENSOR37 型傅里叶红外光谱仪采集其近红外光谱;然后,分别经 SNV 和小波变换方法预处理,结合偏最小二乘法( PLS )建立了棉成分的定量分析模型。结果表明:经 SNV 预处理后的光谱数据结合偏最小二乘法( PLS )建立的模型具有较高的分析精度, 模型的交叉验证均方根误差( RMSEC V) 和预测均方根误差( RMSEP )分别为 和 。参考值和预测值的相关系数为 。基本满足了纺织品领域快速定量检测的精度要求,可以作为纺织品棉含量检测的有效手段。关键词:近红外光谱;纺织品;棉含量;偏最小二乘法;预处理 Rapid d etermination of cotton content in textile by near-infrared spectroscopy technology Abstract In this work, a calibration model for determining the content in textile precisely was established. The near infrared spectra of138 textile samples were collected using TENSOR 37 fourier transform near infrared spectrometer, and SNV and wavelet transform were used to pretreat the spectra. C alibration models for cotton content were established with partial least squares(PLS) method . The experimental results indicate that the model was established by the spectral date after SNV pretreament with partial least squares(PLS) method has higher analysis accuracy. With the model, the Root Mean Square Error Of Cross Validation( RMSECV ) and Root Mean Square Error Of Prediction( RMSEP ) are and , respectively. And a good correlation between reference concentrations and predicted values( R =) was obtained. The proposed approach can basically meet the accuracy requirement of the rapid quantitative detection in the textile field and can appear tobea valid alterative for detection of cotton content in textile. Keywords :near infrared (NIR) spectroscopy;textile;cotton content;Partial Least Square(PLS) method; pretreament 目录摘要............................................................................................................................. I Abstract ...................................................................................................................... II 第一章绪论.............................................................................................................. 1 纺织品行业概述...