电子鼻信号优化在茶叶贮藏时间识别中应用

he electronic nose (e-nose) was applied to predict the storage time of the tea, the volatile components of dry tea leaf, tea beverage and tea remains (the tea leaves is brewed by the water of 100℃ keeping for 3 min and filtered, the filtrate is called tea beverage and the wet tea leaves is called tea remains) were detected by the e-nose, respectively. The first five principal components values was extracted and acted as the inputs of the discrimination analysis in order to decrease the data dimension and optimize the feature vector. The storage time of the Longjing tea was discriminated by the linear discrimination analysis (LDA) and was predicted by the back-propagation neural network (BPNN). The results show that the discrimination and testing results of the dry tea leaf were better than those of tea beverage and tea remains. The predicted error for the storage time by the dry tea leaf was the best except for the fresh tea, and the discrimination result of LDA accord with that of BPNN.