玫瑰香葡萄贮藏期间糖酸品质的近红外检测  

Detection of sugar-acid quality of Muscat grape during storage using NIR spectroscopy

作　　者：陈辰[1] 鲁晓翔[1] 张鹏 陈绍慧 李江阔 

机构地区：[1]天津商业大学生物技术与食品科学学院天津市食品生物技术重点实验室,天津300134 [2]国家农产品保鲜工程技术研究中心天津市农产品采后生理与贮藏保鲜重点实验室,天津300384

出　　处：《食品与发酵工业》2015年第6期175-180,共6页Food and Fermentation Industries

基　　金："十二五"国家科技支撑计划项目(2012BAD38B01);天津市高等学校创新团队培养计划项目(TD12-5049)资助

摘　　要：利用可见/近红外漫反射光谱技术,建立玫瑰香葡萄贮藏期间糖酸品质的快速检测模型。以不同温度(0、10℃)贮藏的玫瑰香葡萄为材料,在全波长光谱内(400~2 500 nm)采用改进偏最小二乘法(MPLS)建立以可溶性固形物(SSC)、总酸(TA)为评价指标的糖酸品质数学模型,讨论光谱采集方法、光谱预处理方法对建模效果的影响,比较不同贮藏温度模型及混合温度糖酸综合模型的适用性。结果显示:MPLS结合一阶导数、反相多元离散校正的处理方法适用于0℃及10℃贮藏果实SSC及10℃贮藏下TA的建模,验证标准差SEP分别为0.289、0.336、0.016,决定系数R2P为0.959 2、0.943 4、0.895 4,相对分析误差RPD为4.23、4.13、3.00;0℃果实的TA模型一阶导数与标准正常化相结合为最优,SEP为0.02,R2P为0.864 3,RPD为2.61。采用最优预处理方法建立的2种温度混合样品糖酸综合模型验证标准差SEP分别为0.390、0.027,决定系数R2P为0.920 5、0.8312,预测效果比单一温度建模略差。表明玫瑰香葡萄果实糖酸品质的近红外模型检测具有可行性,且分温度建模精度较高,综合模型需进一步完善。The aim of this study was to establish near infrared diffuse reflection spectroscopy( NIR) model to detect soluble solids content( SSC),total acid( TA) of Muscat grapes during storage in different temperature( 0℃,10℃). Modified partial least squares( MPLS) in full wavelength spectral range( 400 ~ 2 500 nm) was applied,and the effects on different spectral acquisition and pretreatment methods were discussed. The applications of storage temperature model and comprehensive acid sugar model were compared. The results showed MPLS model combined D1Log( 1 / R) and inverse multivariate discrete correction was suitable for SSC model in all samples and TA model stored at 10 ℃. The validation standard deviation SEP were 0. 289,0. 336,0. 016,respectively; the determination coefficient RP2 were 0. 9592,0. 943 4,0. 895 4,respectively; RPD were 4. 23,4. 13 and 3 respectively; Models with D1Log( 1 / R) and standard normal pretreatment methods of TA stored at 0℃ was the best,SEP was 0. 02,RP2 was 0. 8643,RPD was 2. 61. The SEP of SSC,TA comprehensive model established with the optimal pretreatment methods were0. 390,0. 027,respectively; RP2 were 0. 920 5,0. 831 2. The prediction was not as good as single temperature model.This study indicates that fast and nondestructive detection of sugar and acid in Muscat grapes during storage using visible / near-infrared diffuse reflectance technique is feasible and models including different storage temperature have higher precision prediction. The comprehensive model needs to be further studied to improve it's precision.

关 键 词：可见/近红外漫反射光谱 玫瑰香葡萄 可溶性固形物 总酸 预测模型 

分 类 号：TS255.1[轻工技术与工程—农产品加工及贮藏工程] O657.33[轻工技术与工程—食品科学与工程]