论文部分内容阅读
以298份水稻种子为样本,用常规法测定直链淀粉含量,采用偏最小二乘法(PLS),优化建立精米直链淀粉含量近红外光谱预测校正模型。模型校正决定系数RC为0.95;校正标准差SEC为1.58;内部交叉检验决定系数RP为0.91,标准误差SEP为1.92。利用20个样品进行外部检验,预测值与真实值之间差异不显著,其相关系数达95%以上。定标模型预测性能较好,可以替代化学分析法快速测定水稻直链淀粉含量。
Using 298 rice seeds as sample, the content of amylose was determined by the conventional method. Partial least square method (PLS) was used to optimize the prediction model of amylose content in rice by near infrared spectroscopy. The model calibration coefficient of determination RC is 0.95, the standard deviation of calibration (SEC) is 1.58, the internal cross-validation coefficient RP is 0.91 and the standard error SEP is 1.92. Using 20 samples for external inspection, the difference between the predicted value and the true value is not significant, and the correlation coefficient reaches more than 95%. The calibration model has good predictive performance and can be used as an alternative to chemical analysis for rapid determination of amylose content in rice.