【摘要】 利用近红外反射光谱分析技术和偏最小二乘回归法(PLS),通过比较不同光谱范围和光谱预处理方法,采用二阶导数光谱预处理,在7540.3~5361.1 cm-1和4882.9~4504.9 cm-1谱区内建立了近红外光谱测定玉米秸秆纤维素含量的校正模型。利用15个玉米秸秆样品对所建模型的实际预测效果进行了验证,预测值与化学值的相关系数(r)可达0.9953,最大相对误差仅为5.20。结果表明,近红外光谱技术可以快速、准确地测定玉米秸秆纤维素,该结果对玉米秸秆材料的快速鉴定和筛选利用具有重要的意义。更多还原

【Abstract】 The near infrared reflectance spectroscopy（NIRS） and partial least square regression（PLS） were used to establish the models by comparing several preprocessing procedures and wavelength ranges.The（optimal） models could be obtained in the range of 6101.7⁵773.8 cm^-1 and 4601.3～4246.5 cm^-1 by the spectral data preprocessing of the second derivative.In addition,15 additional maize stalk samples were used to evaluate the reliability of the calibration model.The corelation coefficient was 0.9953 between NIRS（predicted） and actual cellulose in these materials.Thus,the accuracy of prediction could be comparable to that of chemical method.The results show that NIRS is a simple effective means for measuring cellulose in maize stalk.The results are of great important in screening and evaluating quality constituents of silage maize.更多还原