电子鼻-主成分分析-线性回归拟合法检测江平虾冷藏过程中的新鲜度

摘要

The freshness of Jiangping shrimp stored under refrigeration was qualitatively and quantitatively analyzed by electronic nose-principal component analysis-linear regression fitting model (EN-PCA-LRFM). Jiangping shrimp with different refrigeration times were detected with an electronic nose, the response values to volatile components were collected for qualitative and quantitative analysis by PCA and LREM, respectively. Lastly, an EN-PCA-LRFM method was presented to evaluate the freshness of refrigerated shrimp as a function of storage time. The results showed that sensor S1, S3, S4, S5 and S6 in the electronic nose gave better response to the samples in eight groups, and the mean response values between 45 and 49 s were selected as eigenvalues. PCA allowed complete discrimination of the samples with different refrigeration times and therefore could be used for qualitative analysis. The correlation coefficient of the developed LREM model was 0.8841, and there was a linear relationship between the predicted and actual values, which showed that this model had good generality. Therefore, this study demonstrated the viability of using the EN-PCA-LRFM method to qualitatively and quantitatively analyze the freshness of Jiangping shrimp.%利用电子鼻-主成分分析-线性回归拟合模型(electronic nose-principal component analysis-linear regression fitting model,EN-PCA-LRFM)法对江平虾冷藏过程中的新鲜度进行定性和定量分析.对不同冷藏时间的江平虾进行EN检测,记录对应挥发性成分的响应值,然后利用PCA和LRFM分别进行定性判别和定量分析,构建检测江平虾冷藏过程中新鲜度变化的EN-PCA-LRFM法.结果表明:电子鼻的S1、S3、S4、S5、S6传感器对8组样品的响应较好,45~49 s内的平均响应值可作为特征值进行分析;PCA能够对不同冷藏时间的样品进行明显区分,可用于定性分析;LRFM的决定系数为0.8841,预测值和真实值呈现一定的线性关系,表明该模型具有较好的泛化性.基于EN-PCA-LRFM法定性和定量检测江平虾冷藏过程中的新鲜度是可行的.