文献类型：期刊文献

中文题名：近红外光谱信号的优化处理

英文题名：Optimization method of near infrared spectroscopy signal

作者：李冬云[1];李彩云[2]

第一作者：李冬云

机构：[1]北京联合大学信息学院;[2]平原大学化学与环境工程学院

第一机构：北京联合大学智慧城市学院

年份：2007

卷号：24

期号：10

起止页码：1418-1420

中文期刊名：计算机与应用化学

外文期刊名：Computers and Applied Chemistry

收录：CSTPCD;;北大核心:【北大核心2004】；CSCD:【CSCD2011_2012】；

语种：中文

中文关键词：近红外光谱;主成分数;建模谱区;求导;平滑处理

外文关键词：near infrared spectroscopy, number of principal component, range of wavelength, differentiation, smoothing

摘要：用实验室常规方法测定竹材样品的木质素含量,漫反射方式采集样品的近红外光谱信号,偏最小二乘法(PLS)和完全交互验证方式以建立毛竹木质素含量的定量分析模型。研究主成分数、建模谱区、求导和平滑预处理技术对定量分析模型的影响。结果表明,预处理技术压缩和恢复的近红外光谱信号效果良好,提高了模型的预测能力,优化近红外定量分析模型有重要参考价值。
The quantitative analysis model for lignin content in bamboo was established using the partial least square （PLS） regression and the full cross validation technique. The lignin contents of bamboo samples were analyzed according to traditional chemical method, the spectra of which were collected by near infrared reflectance spectroscopy （NIRS）. The effects of number of principal component, range of wavelength, differentiation and smoothing of raw spectra on model were investigated. The result showed that the pretreatment of spectra was beneficial to data compression and signal reconstruction, resulting in the improvement of prediction accuracy.