论文部分内容阅读
应用傅立叶变换近红外光谱技术,建立锅盔水分含量分析模型。测定61份锅盔的近红外光谱,经一阶导数+MSC预处理以滤去噪声,在7 501.9~4 597.6 cm-1谱段范围内,选择维数10,利用偏最小二乘法建立近红外光谱与国标参考方法测得的水分含量之间的相关模型。最终得到水分定量校正模型决定系数(R2)为99.03%,内部交叉验证均方差(RMSECV)为0.409%。用该模型对19个未知锅盔样品进行外部验证,其水分外部验证决定系数(R2)为97.99%,预测标准偏差(RMSEP)为0.341%。结果表明,近红外定量分析技术有较高的准确度,能满足锅盔水分的快速检测精度要求。
Fourier transform near-infrared spectroscopy was used to establish the model of moisture content analysis. The near infra-red spectrum of 61 helmets was determined. The first derivative + MSC was pretreated to filter out noise. The dimension of 7 501.9 ~ 4 597.6 cm-1 was chosen as the dimension 10 and the near-infrared spectrum was established by partial least squares And the national standard reference method to measure the moisture content of the correlation model. The results showed that the coefficient of determination (R2) of quantitative calibration model was 99.03% and the mean square error of internal cross validation (RMSECV) was 0.409%. The model was used to verify the exterior of 19 unknown samples of pots. The determination coefficient (R2) of water exterior validation was 97.99% and the prediction standard deviation (RMSEP) was 0.341%. The results show that NIR quantitative analysis technology has high accuracy and can meet the requirement of rapid detection accuracy of the water in the helmet.