摘要
探讨了快速、无损检测食醋中总酸含量的建模方法,利用近红外光谱法分别结合间隔偏最小二乘法(iPLS)、反向区间偏最小二乘法(BiPLS)、联合间隔偏最小二乘算法(SiPLS)进行建模,对各算法在不同划分区间数及区间选择时对建立模型的影响进行比较。结果表明:BiPLS、SiPLS(2,3,4区间联合)建模效果较好于iPLS所建立的模型,其中BiPLS在选择43个子区间,5个子区间联合(3,4,6,7,16)最佳,其RMSECV和RMSEP分别为0.2876和0.2726,校正集和预测集相关系数分别为0.9343和0.938;SiPLS在选择3个区间联合,49个区间数(3、5、7区间联合)最佳,其RMSECV和RMSEP分别为0.2607和0.2802,校正集和预测集相关系数分别为0.9463和0.9371;iPLS在选择22个子区间,第三个子区间,主因子数为4时最佳,其RMSECV和RMSEP分别为0.2998和0.2977,校正集和预测集相关系数分别为0.928和0.9213。不同偏最小二乘算法所选取区域大多集中于5500~6000 cm-1范围内,证明该波数范围应该是总酸的相应特征区间。
Rapid non-destructive testing of total acid content in vinegar Modeling.Using near infrared spectroscopy were combined with interval Partial Least Squares(iPLS),Backward interval Partial Last Squares(BiPLS),Synergy interval Partial Least Squares algorithm(SiPLS)to modeling,division of each algorithm in different intervals and the interval chosen to model the impact when compared.The results show that the BiPLS,SiPLS(2,3,4 joint interval) model is better on the model iPLSi,43 of them in the choice of sub-interval BiPLS,5 sub-range joint(3,4,6,7,16) best,the RMSECV and RMSEP were 0.2876 and 0.2726,calibration and the prediction correlation coefficient of 0.9343 and 0.938;SiPLS joint in the choice of three intervals,49 intervals(3,5,7 interval joint) best,the RMSECV and RMSEP were 0.2607 and 0.2802,calibration and the prediction correlation coefficient of
出处
《中国调味品》
CAS
北大核心
2011年第1期107-110,113,共5页
China Condiment
基金
全国优秀博士基金(200968)
国家博士后基金(20070411024)
江苏大学拔尖人才启动基金
关键词
醋
近红外
偏最小二乘算法
vinegar
near infrared
Partial Least Squares