利用核主成分分析(KPCA)对苹果近红外光谱进行特征提取,不但使得光谱维数大幅降低,而且能有效地提取原始光谱的非线性信息。实验表明,KPCA结合支持向量回归机(SVR)建立的苹果酸度回归模型与PCA-SVR和SVR模型相比,提高了预测精度,缩短了训练时间和预测时间,是一种有效的光谱特征提取方法。 The kernel principal component analysis (KPCA) of the apple near infrared spectroscopy feature extraction, not only significantly reduce the spectral dimension, but also effectively extract the original spectral nonlinear information. Experiments show that compared with PCA-SVR and SVR model, KPCA combined with Support Vector Regression (SVR) can improve the prediction accuracy and shorten the training time and prediction time, which is an effective spectral feature extraction method .