近红外光谱法测定毛竹综纤维素的含量研究

研究了用近红外光谱(NIR)结合多变量统计分析技术对毛竹综纤维素含量的快速测定。用常规实验室方法测定了54个竹材样品的综纤维素含量,用近红外光谱仪采集相应样品的光谱,对原始光谱进行二阶导数和25点平滑预处理后,从54个竹材样品中挑选41个代表性的样品建模,选择1011~1675nm和1930~2488nm波段区间,用偏最小二乘法(PLS1)和完全交互验证方式建立毛竹综纤维素含量的预测模型。结果表明,毛竹综纤维素含量和近红外光谱之间存在非常好的相关性,预测模型的相关系数(RP)为0.95,预测模型的标准偏差(SEP)为0.76%。

Study on Determination of Holocellulose Content of Bamboo by Near-infrared Reflectance Spectroscopy

A rapid determination of holocellulose content of bamboo (Phyllestachys heterocycla (Carr.)Mitford cv.pubescens) has been developed by near-infrared reflectance spectroscopy (NIR)coupled by multivariate statistical analysis technique. The holocellulose contents of 54 samples were analyzed with traditional wet chemical technique, the spectra of which were collected by NIR. After the spectra were converted to the second derivative spectra with a gap of 25 nm, the prediction model for the holocellulose content of bamboo samples was established using partial least-square regression (PLS1)and full cross validation in the range of 1011KG-*2]-KG-*6]1675 nm and 1930KG-*2]-KG-*6]2488 nm. High correlations were obtained between the predicted NIR results and those obtained from traditional chemical method. The correlation coefficient of prediction model is 0.95 and standard error of prediction (SEP) is 0.76 %.