Discrimination of Storage Age for Wheat by E-Nose

An investigation was conducted to evaluate the capacity of an electronic nose (E-nose) to classify wheat based on storage age. The effects of vial volume, headspace generation time, and wheat sample mass were studied, and the optimum experimental procedure was determined using multifactor analysis of variance. The results showed that the headspace generation time and the mass of the wheat sample had an obvious effect on E-nose response; however, the volume of the vial had no significant effect on E-nose response. The optimum parameters are 500 mL vial volume, 1.5 h headspace generation time, and 50 g sample mass. The five wheat groups were discriminated completely by principal component analysis (PCA). However, samples from 2006 and 2005 were overlapped by linear discriminant analysis (LDA). The method of artificial neural network (ANN) was performed, and 85% of the testing set was classified correctly using a back-propagation neural network.