电子鼻技术在食醋识别中的应用

利用由10个掺杂纳米氧化锌厚膜气敏传感器组成的阵列对9种食醋和乙酸溶液进行了测量.并通过主元分析、聚类分析和概率神经网络对数据进行了分析和识别.主元分析表明不同的食醋在品牌、种类、酸度等方面具有一定的相似性.聚类分析进一步研究了食醋种类之间的相似程度.利用概率神经网络对所测试的食醋进行了识别,有较高的识别率.分析表明电子鼻技术是食醋分析和识别的一种具有发展前途的实用技术.

Research on vinegars identification by electronic nose

Nine kinds of vinegars and an acetic acid resolution were measured by the gas sensors array which were composed of ten doped nano-ZnO thick film sensors. Principal Component Analysis (PCA), Cluster Analysis (CA) and Probabilistic Neural Network (PNN) were used in the data analysis and pattern recognition. The vinegars could be identified according to their brands, kinds, and the total acidity of the vinegars indicated by the Principal Component Analysis. Cluster Analysis reflected the similarity of the vinegars. Finally, Probabilistic Neural Network was employed to identify the vinegars, and the accuracy of PNN in term of predicting the vinegars was very high. This work shows the potential applications of the electronic nose for analyzing and identifying the vinegars.