| 摘要: |
| 【目的】采用近红外光谱(Near-infrared spectroscopy,NIRS)分析技术对食醋中的总酸和还原糖进行快速定量分析，为食醋品质的检测提供技术支持。【方法】用近红外透射光谱法采集食醋的光谱信息，用Savitzky-Golay一阶导数和小波导数（Wavelet derivative,WD）对采集到的食醋光谱进行预处理，采用连续投影算法（Successive projections algorithm,SPA）的波长筛选方法对食醋的总酸以及还原糖定量分析模型进行优化。【结果】食醋光谱经小波导数预处理和基于SPA的波长优选后，总酸和还原糖含量分析精度得以提高，模型得到简化。总酸的预测均方根误差（Root mean square error of prediction,RMSEP）由0.282降低为0.227，预测相关系数（Correlation coefficient of prediction，Rp）由0.965提高到0.977，建模变量由680个精简为51个；还原糖的RMSEP由0.627降低为0.606，预测相关系数Rp由0.996提高至0.997，建模变量精简为50个。【结论】利用近红外光谱分析技术结合小波导数和SPA分析方法，可用于食醋品质的定量分析，为食醋品质的快速检测提供了一种可行的方法。 |
| 关键词: 食醋 品质检测 总酸 还原糖 小波导数 连续投影算法 近红外光谱 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目（21265006）;江西省青年科学基金项目（20114BAB2011010）;华东交通大学博士启动基金项目（01309021）;江苏省企业博士集聚计划项目 |
|
| Model optimization of near-infrared spectroscopy for determination of total acid and reducing sugar in vinegar |
|
|
| Abstract: |
| 【Objective】Near-infrared spectroscopy (NIRS) was used for the determination of total acid and reducing sugar in vinegar samples.It can provide technical support for vinegar quality detection.【Method】Near-infrared transmittance spectroscopy was adopted for spectral acquisition of vinegar samples.Savitzky-Golay 1st derivative and wavelet derivative were used for spectral pretreatment.Wavelength selection method of successive projections algorithm (SPA) was used for model optimization of total acid and reducing sugar in vinegar samples.【Result】The results showed that the wavelet derivative (WD) combined with SPA could not only improve the analysis precision,but also simplify the modeling variables.For the total acid model,the root mean square error of prediction (RMSEP) and modeling variables were reduced form 0.282 and 680 to 0.227 and 51,respectively,and the correlation coefficient of prediction (Rp) was improved from 0.965 to 0.977.For the reducing sugar model,the RMSEP and modeling variables were reduced form 0.627 and 680 to 0.606 and 50,respectively,and the Rp was improved from 0.996 to 0.997.【Conclusion】NIRS combined with WD-SPA method is a feasible and fast method for the determination of vinegar quality. |
| Key words: vinegar quality determination total acid reducing sugar wavelet derivative successive projections algorithm (SPA) near-infrared spectroscopy (NIRS) |
