【摘要】 采用电子鼻对五个储藏年限的陈化小麦进行年限分析,确定了采用电子鼻判别小麦储藏年限的最佳参数及方法.对传感器信号进行多因素方差分析可知:对于固定容器的陈化小麦样品,不同的小麦密封时间对电子鼻的响应信号的影响极为显著;其次是小麦在烧杯内的密封质量.通过静置密封时间和密封质量的方差分析,得出小麦在500 mL烧杯内的最佳静置时间为1.5 h,密封在烧杯内的小麦最恰当质量为50 g.采用以上参数,对五个储藏年限的小麦进行辨别,PCA分析可以将不同储藏年限的小麦较好的区分开来,并且五个年份的小麦自右上角至左下角依次排列;而LDA分析能够将差别年限较大的陈化小麦进行区分,差距较小的,不能够很好的区分,其区分效果不如PCA分析;进而采用BP神经网络的方法进行判别分析,训练样本正确率为100%,测试样本正确率也达到了85%以上.更多还原

【Abstract】 An investigation was made to evaluate the capacity of an electronic nose(E-nose,Pen 2) to classify the wheat groups with different storage times(W1,W2,W3,W4 and W5).In the experiment,the volume of vial,the headspace generated time and the mass of the wheat sample were considered,and the optimum experimental procedure was determined by using the multiple-factor analysis of variance(ANOVA).The results showed that the headspace generated time had more significantly effect on the result of E-nose response than the mass of the wheat sample with settled volume of vial.Based on ANOVA of the results with the headspace generated time and the mass of the wheat sample,the results showed that the optimum time and mass of the wheat sealed in 500 mL beaker was 1.5 h and 50 g respectively.The five wheat groups were measured and response values at 60 s were conducted by PCA,linear discrimination analysis(LDA) and artificial neural network(ANN).The five wheat groups were discriminated completely by PCA;however,samples of W1 and W2 were overlapped by LDA;100% of the training set and more than 82.5% of the testing set were classified by ANN(network topology 30-12-4).更多还原