Evaluation for fresh juice of strawberries with different storage period using electronic tongue

In order to discuss the change rules of fresh juice quality of strawberry with different cold storage time, the electronic tongue was used to evaluate the physicochemical properties and qualities of fresh juice of strawberries quantitatively and qualitatively. Based on principal component analysis (PCA) and linear discriminant analysis (LDA), it was concluded that fresh juice of strawberries stored for different time could be described by electronic tongue well. In early storage of strawberries, quality of the fresh juice decreased rapidly. With middle storage period of strawberries, quality of the fresh juice decreased slowly. Back propagation neural network (BPNN) was also employed to predict the storage time of the samples. The result show that BPNN can get 94.9% training accuracy and 85.71% prediction accuracy. Partial least-square (PLS) regression and diverse line regression were employed to predict the component of fresh juice. PLS regression was more capable of building models with good prediction accuracy (R2>0.9) than multiple line regression, which showed the good predictive validity on the quality of fresh juice of strawberries.