标题:Design and application of electronic tongue system for orange juice quality detection using internet of things
作者:Ma, Zeliang; Yin, Tingjia; Gui, Tingting; Wang, Zhiqiang; Sun, Xia; Li, Caihong; Guo, Yemin
通讯作者:Wang, ZQ
作者机构:[Ma, Zeliang; Yin, Tingjia; Gui, Tingting; Wang, Zhiqiang; Li, Caihong] Shandong Univ Technol, Coll Comp Sci & Technol, Zibo 255049, Peoples R China.; 更多
会议名称:6th International-Federation-of-Automatic-Control (IFAC) Conference on Bio-Robotics (BIOROBOTICS)
会议日期:JUL 13-15, 2018
来源:IFAC PAPERSONLINE
出版年:2018
卷:51
期:17
页码:437-442
DOI:10.1016/j.ifacol.2018.08.182
关键词:internet of things; electronic tongue; pattern recognition; orange juice; quality
摘要:Imitating the mechanism of human taste perception, an electronic tongue (ET) system for orange juice quality detection based on internet of things was developed. The system consists of three parts: hand-held detection telininal, wireless communication system and cloud service platfoun. During the detection, the hand-held detection telininal was used to obtain taste "fingerprint" data from various kinds of juices based on large pulse scanning potential, and then transmitted these data to a cloud service platfoun through a wireless communication system, as well as the pattern recognition methods were utilized to analyze the data. Finally, the results were compared with the built-in "fingerprint" database of the beverage on the cloud platfoun so as to obtain the brand or adulteration infounation of the tested juice. In this study, the developed ET system was further used for the identification of orange juice brand and purity detection. Linear discriminant analysis (LDA) was applied to the qualitative analysis of orange juice with different brand, and the support vector machine (SVM) was used to quantitative forecast of orange juice purity. The result indicated that the LDA can effectively differentiate samples of orange juice with different brand, which the identification accuracy reached 100%; the SVM model possess a good quantitative prediction accuracy for orange juice purity with the root mean square error (RMSE) in prediction set reaching 0.0172. This system possessed the advantages of fast detection, simple operation, low cost, high stability as well as the result was easy to query. (C) 2018, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.
收录类别:CPCI-S;EI
资源类型:会议论文;期刊论文
TOP