为了实现食醋品种的准确分类,探索应用电子鼻技术和两种特征提取方法进行食醋的检测和分类。先用自制电子鼻系统检测5个品种食醋的电子鼻信号,接着用标准正态变量变换进行数据预处理,然后分别用主成分分析(principal component analysis, PCA)+线性判别分析(linear discriminant analysis, LDA)和正交线性判别分析(orthogonal linear discriminant analysis, OLDA)对食醋电子鼻信号进行降维与特征提取,最后用最近邻分类器进行分类。实验表明,PCA+LDA的分类准确率最高达到90.32%,而OLDA的分类准确率最高达到91.52%。另外,PCA+LDA需要2次特征提取而OLDA只要1次。因此,OLDA在特征提取方面要优于PCA+LDA,基于OLDA和电子鼻技术的食醋品种分类方法是切实可行的。
In order to classify vinegar varieties correctly, the electronic nose (E-nose) technology was explored in the application of two feature extraction methods to detect and classify vinegars. After the detection of E-nose signals of 5 brands of vinegars using our designed E-nose system, the signals were pretreated with the standard normal variate transformation. Then principal component analysis (PCA) plus linear discriminant analysis (LDA) and the orthogonal linear discriminant analysis (OLDA) were introduced to reduce the dimension and extract the features of the E-nose signals of vinegars. At last, nearest neighbor classifier was used to classify the data. The results showed that the highest classification accuracy of PCA+LDA was 90.32%, while that of OLDA was 91.52%. On the other hand, PCA+LDA required twice feature extractions, while OLDA needed only once. Therefore, OLDA is superior to PCA+LDA in feature extraction, and it is a feasible method to use OLDA coupled with E-nose technology for the classification of vinegar varieties.
[1] 魏泉增,汤雅,李伟民.液相色谱-串联质谱测定食醋中生物胺含量[J].中国酿造,2019,38(4):170-173.
[2] 单伟国.基于专家系统的挥发性气体检测的电子鼻研究[D].合肥:中国科学技术大学,2010.
[3] HASSAN M,BERMAK A.Robust bayesian inference for gas identification in electronic nose applications by using random matrix theory[J]. IEEE Sensors Journal, 2016, 16(7):2 036-2 045.
[4] ZHANG Lei,TIAN Fengchun,NIE Hong,et al.Classification of multiple indoor air contaminants by an electronic nose and a hybrid support vector machine[J].Sensors and Actuators B: Chemical, 2012,174:114-125.
[5] ZHANG Lei,LIU Yan,DENG Pingling.Odor recognition in multiple E-nose systems with cross-domain discriminative subspace learning[J].IEEE Transactions on Instrumentation and Measurement,2017,66(7):1-14.
[6] 陈芳,黄玉坤,苑阳阳,等.基于电子鼻无损鉴别掺假蜂蜜[J].西华大学学报(自然科学版),2018,37(5):56-60.
[7] 周博,嵇云,蔡国华,等.电子鼻检测农作物病虫害的研究进展[J].江苏农业科学,2019,47(15):143-148.
[8] 孟洁,商细彬,荆博宇,等.污水处理厂还原硫化物和臭气浓度的电子鼻监测技术探究[J].环境监测管理与技术,2019,31(3):45-48.
[9] 方向生,施汉昌,何苗,等.电子鼻在环境监测中的应用与进展[J].环境科学与技术,2011,34(10):112-117;122.
[10] 冷晓红,陈海燕,郭鸿雁.电子鼻技术在中药领域的应用[J].西北药学杂志,2019,34(3):426-428.
[11] 杨瑞琦.基于电子鼻的易霉变中药质量快速评价[D].北京:北京中医药大学,2019.
[12] 田连起,乐智勇,曹晖,等.基于电子鼻技术的中药蕲蛇饮片炮制矫味物质基础研究[J].中医学报,2019,34(4):785-789.
[13] 费程浩,戴辉,苏杭,等.电子鼻技术的研究进展及其在中药行业中的应用[J].世界中医药,2019,14(2):257-262.
[14] 张井,张维一,李燕,等.基于电子鼻分析的碳酸饮料识别判定技术研究[J].食品与发酵工业,2017,43(5):214-218.
[15] 郑舒文,陈卫华.基于电子鼻和电子舌技术的鳕鱼鲜度评定[J].中国调味品,2019,44(5):164-169.
[16] 杨潇,郭登峰,王祖文,等.基于电子鼻的猪肉冷冻储藏期的无损检测方法[J].食品与发酵工业,2018,44(3):247-252.
[17] 张四喆,贾文珅,马洁,等.一种高效的冷鲜肉新鲜度检测工具—电子鼻[J].分析试验室,2019,38(7):878-884.
[18] 马会杰,蒋宾,潘玉兰,等.电子鼻对不同产地名优绿茶和红茶香气特征的判别研究[J].食品科技,2019,44(1):336-344.
[19] WU Xiaohong,ZHU Jin,WU Bin,et al.Discrimination of Chinese liquors based on electronic nose and fuzzy discriminant principal component analysis[J].Foods,2019,8(1),38.
[20] 成剑峰,胡红娟.食醋半成品贮存期质量监测方法的探讨[J].中国酿造,2018,37(11):102-105.
[21] 黎新荣.电子鼻在沃柑贮藏时间识别中的应用[J].南方农业学报,2018,49(9):1 827-1 832.
[22] 魏永义,焦驼文,崔广,等. 食醋特征香气成分分析研究[J].中国调味品,2010,35(6):99-101.
[23] 嵇港.基于电子鼻系统的食醋分类研究[D].镇江:江苏大学,2018.
[24] YE JIEPING.Characterization of a family of algorithms for generalized discriminant analysis on undersampled problems[J]. Journal of Machine Learning Research,2005,6(1):483-502.
[25] KASBE M S,DESHMUKH S L,MUJAWAR T H,et al.An electronic nose with LabVIEW using SnO2 based gas sensors: application to test freshness of the fruits[J].International Journal of Scientific & Engineering Research,2015,6(4):1 977-1 982.
[26] 刘鹏,叶宾.基于随机矩阵理论的高维数据线性判别分析方法[J].计算机科学,2019,46(S1):423-426.
