采用近红外光谱(NIRS)分析技术和化学计量方法建立小麦脂肪酸值的近红外分析模型，并对模型进行预测准确性评价。结果表明：所建立模型的定标相关系数（RSQ）为0.9026，交叉验证相关系数（1-VR）为0.6278，定标标准偏差（SEC）为3.8735，交叉验证标准偏差（SECV）为7.0908。外部验证的相关系数(r)为0.948，外部验证标准偏差(SEP)为3.8709。标准方法与NIRS测定方法测定的小麦脂肪酸值含量之间的t 检验值为1.345，显示两种方法测定结果无显著性差异(P ＜ 0.05)，预测值与实测值的平均绝对偏差为0.25，说明所建立的稻谷脂肪酸值的NIRS 数学模型预测准确性较好，可用于小麦脂肪酸值的快速预测。

The mathematic models for the prediction of fatty acid content of wheat were established with the technique of near infrared spectroscopy (NIRS). The result showed that the coefficient of correlation (RSQ) was of 0.9026 and the standard error of calibration (SEC) of 3.8735. The determination coefficient of cross-validation (1-VR) was of 0.6278 and the standard error of cross-validation (SECV) was of 7.0980. Regression squared (r) was of 0.948 and square error of prediction (SEP) was of 3.8709. The t test value between the chemical standard methods and NIRS method was 1.345 suggesting no significant difference between these two methods. The absolute average deviation was 0.25. This NIRS method could be applied to predict the fatty acid content in wheat.

河南工业大学高层次人才基金，国家十二五科技计划课题（2012AA101705）