基于MC-UVE、GA算法及因子分析对葡萄酒酒精度近红外定量模型的优化研究
首发时间:2017-12-12
摘要:本实验主要对葡萄酒酒精度偏最小二乘(Partial least squares, PLS)回归模型进行优化研究。使用近红外光谱仪采集葡萄酒样本的光谱数据,用于建立酒精度定量模型,实现在线快速检测。使用蒙特卡罗无信息变量消除(MC-UVE)和遗传算法(GA)进行变量选择,基于被选择的变量分别进行PLS和因子分析(Factor analysis, FA)建立回归模型。结果表明,MC-UVE-GA-FAR模型预测集相关系数为0.946、预测均方根误差为0.215,效果优于MC-UVE-GA-PLS模型。虽然与基于全范围原始光谱所建PLS回归模型相比,模型效果略有下降,但是模型所选变量个数仅为6,极大地简化了模型,说明MC-UVE和GA算法可以实现模型的优化。
关键词: 近红外光谱 葡萄酒 遗传算法 蒙特卡罗无信息变量消除 因子分析
For information in English, please click here
Optimization of Near Infrared Quantitative Model for Wine Alcohol Content Based on MC-UVE, GA Algorithm and Factor
Abstract:This study is focused on the optimization of the partial least squares (PLS) regression model of wine alcohol content. The near-infrared spectroscopy was used to collect the spectral data of the wine samples and the data was used to establish the quantitative model of alcohol to achieve rapid on-line detection. Partial least squares (PLS) regression model and factor analysis(FA) model were established based on the selected variables, chosen by Monte-Carlo uninformative variable elimination (MC-UVE) and genetic algorithm (GA). The results showed that the MC-UVE-GA-RAR model, which yielded correlation coefficient of 0.946 and root mean square error of prediction of 0.215, was superior to the MV-UVE-GA-PLS model. In comparison of the performances of the full-spectra PLS regression model, the model based on the selected wave numbers were slightly lower, but 6 variables in total were selected, which greatly simplified the model. The study indicates the MC-UVE and GA algorithms can optimize the model.
Keywords: Near-infrared spectroscopy Wine Genetic algorithm Monte-Carlo uninformative variable elimination Factor analysis
引用
No.****
同行评议
共计0人参与
勘误表
基于MC-UVE、GA算法及因子分析对葡萄酒酒精度近红外定量模型的优化研究
评论
全部评论0/1000