不同成熟度草莓鲜榨果汁的电子鼻和电子舌检测

Evaluation of fresh juice of strawberries at different degrees of ripeness using electronic nose and electronic tongue

采用电子鼻和电子舌,通过主成分分析(principal component analysis,PCA)、线性判别分析(linear discriminantanalysis,LDA)对不同成熟度草莓鲜榨果汁的风味品质进行检测区分,并通过偏最小二乘回归分析(partial leastsquares,PLS)建立电子鼻和电子舌传感器响应信号与草莓鲜榨汁理化指标之间的关系,定量预测草莓的品质指标.结果表明:PCA和LDA法均对不同成熟度草莓鲜榨汁的区分效果较好,且电子舌的区分效果好于电子鼻,电子鼻和电子舌传感器响应信号融合后的区分效果与电子舌相当;通过PLS法,电子舌传感器响应信号能够较好地预测不同成熟度草莓鲜榨汁的可溶性固形物含量、Vc含量和pH值,但对总酸含量的预测能力稍差,其校正决定系数R2为0.876,预测决定系数R2为0.793;电子鼻和电子舌传感器响应信号融合后能够很好地预测不同成熟度草莓鲜榨汁的可溶性固形物、Vc和总酸含量及pH值,其预测能力好于单一的电子鼻或电子舌,其中对于电子舌不能很好预测的总酸含量其校正决定系数和预测决定系数分别上升为0.965和0.908.表明采用上述方法能对样品进行较好地区分且能对样品的品质指标进行较好地预测,其中电子鼻和电子舌信号融合对样品的区分能力和预测能力增强,验证了电子鼻和电子舌结合是对样品气味和味道的综合信息进行评价.

Fresh fruit juice has become more and more popular, since it has high nutritional value, thus, more attention is paid to its quality. However, traditional test methods are often time-consuming and cumbersome, can not realize real-time detection and obtain accurate and overall information only with single test results. Electronic nose （e-nose） and electronic tongue （e-tongue）, which simulate human nose and tongue system, are new rapid detection instruments. It is found that both e-nose and e-tongue system can achieve rapid detection to get the overall information, with high sensitivity, reliability and repeatability. At present, research in fresh fruit juice using e-nose and e-tongue is still in the preliminary stage, and they have not been found in the fresh strawberry juice evaluation. In this study, an e-nose and an e-tongue were employed to detect the fresh juice of strawberries at different degrees of ripeness. Data were analyzed with three pattern recognition techniques： principal component analysis （PCA）, linear discriminant analysis （LDA） and partial least square （PLS）. The results showed that the fresh juice of strawberries at different degrees of ripeness could well be discriminated by PCA and LDA, and the result from e-tongue and e-nose was proved to be superior to that of e-nose only for classifying the samples. And the combination of e-tongue and e-nose had the similar level compared with e-tongue （Fig. 7-9）. PLS was used to establish the relationship between response signals of e-nose and e-tongue and physicochemical indexes of fresh strawberry juice. Analysis of PLS showed that response signals of e-tongue could predict the value of soluble solids content, Vc content and pH value, but total acid content was recorded with the correction and prediction coefficient （R2） of determination 0.876 and 0.793, respectively （Fig. 10）. However, the combination of e-nose and e-tongue with PLS could more efficiently predict soluble solids content, Vc content, total acid content and pH value （Fig. 11）, among which the correction and prediction coefficient of determination of total acid content increased to 0.965 and 0.908, respectively. In conclusion, the combination of e-nose and e-tongue reveals better abilities to discriminate and predict fresh juice, which is confirmed to be a general detection for the comprehensive smell and taste information.