嗅觉可视化技术在白酒识别中的应用

Chinese Liquors Identification by Olfaction Visualization Technology

研制了由30个可视化传感器组成的可视化传感器阵列,对4种白酒进行了测试。通过主成分分析、聚类分析和BP神经网络对实验数据进行了分析和识别。主成分分析表明可视化传感器阵列不仅能够很好地区分不同酒精度的白酒,而且也能区分酒精度相似、香型不同的白酒。聚类分析进一步验证了主成分分析的结果,但部分样本不能够用聚类分析区分。利用BP神经网络对测试样本识别,可以完全区分4种测试白酒。

Four kinds of Chinese liquors were detected by the colorimetric sensor array, which was composed of thirty chemoresponsive sensors. Principal component analysis （PCA）, cluster analysis （CA） and back-propagation artificial neural network （BP-ANN） were used in the data analysis and pattern recognition. With principal component analysis, not only Chinese liquors could be identified according to alcoholicity, but also bouquet. Cluster analysis further proved the result of PCA, but some samples could not be classified by using this method. Finally, BP neural network was employed to identify the Chinese liquors, and the accuracy of recognition was 100%. This research shows the potential applications of the olfaction visualization technology for analyzing and identifying Chinese liquors.