摘要
为了实现酱油品种的准确识别,先使用电子鼻系统对信号进行采集,再用主成分分析(PCI)和线性判别分析(LDA)对数据进行降维和特征提取,之后使用改进的GK算法(M_GK)进行聚类,并与模糊C均值算法(FCM)和传统GK聚类进行比较;实验结果表明M_GK聚类的准确率高于FCM和GK聚类。因此电子鼻技术结合M_GK聚类可以实现酱油品种的准确识别,提供了实现酱油分析鉴别的一种有效模型。
In order to realize the accurate identification of soy sauce varieties,firstly,the electronic nose system was used to collect the signals,then the dimensionality reduction and feature extraction of the data were carried out by principal component analysis(PCI)and linear discriminant analysis(LDA).Then,the modified GK algorithm(M_GK)was used for clustering,which was compared with fuzzy c-means algorithm(FCM)and traditional GK clustering.The experimental results showed that the accuracy of M_GK clustering was higher than that of FCM and GK clustering.Therefore,electronic nose technology combined with M_GK clustering could realize the accurate identification of soy sauce varieties and provided an effective model for soy sauce analysis and identification.
作者
钮永莉
邹长忠
贾红雯
NIU Yongli;ZOU Changzhong;JIA Hongwen(ChuZhou Vocational Technology College,Chuzhou,Anhui 239000;Fuzhou University,Fuzhou,Fujian 350002,China)
出处
《九江学院学报（自然科学版）》
CAS
2022年第3期37-40,共4页
Journal of Jiujiang University：Natural Science Edition
基金
2020安徽省质量工程项目(编号2020kfkc370)
滁州职业技术学院校级科研重点项目(编号YJZ-2020-12)
滁州职业技术学院校级人才项目(编号ZD2019015)的成果之一。
