利用主成分-所有可能回归法，建立了烤烟、小麦样品不同组份的近红外光谱定量分析模型。结果表明，烤烟样品的总糖、还原糖以及小麦样品的蛋白质含量的预测模型均有好的定量分析结果，且其预测结果与PLS法预测结果相当。 Using principal components - all possible regression methods, a quantitative analysis model of near infrared spectra of different components of flue-cured tobacco and wheat samples was established. The results showed that the prediction models of protein content of total sugar, reducing sugar and wheat samples of flue-cured tobacco samples had good quantitative analysis results, and the prediction results were in good agreement with the PLS method.