Dissertation
Dissertation > Agricultural Sciences > Livestock, animal medicine,hunting,silkworm,bee > Beekeeping,beneficial insects feeding > Bee products processing and utilization > Honey

The Identification of Honey Based on the Voltammetric Electronic Tongue

Author ZhangYanPing
Tutor MenHong
School Tohoku Electric Power University
Course Control Science and Engineering
Keywords voltammetric electronic tongue honey PCA FKNN Fuzzy ARTMAP
CLC S896.1
Type Master's thesis
Year 2014
Downloads 13
Quotes 0
Download Dissertation

Honey, as a kind of nutrition and health care products, not only attracts more andmore people’s attention, but also has many advantages, like refreshing taste,appropriate sticky and so on. In recent years, honey adulteration has repeatedlyoccurred, and the quality of honey is various on the market, so the adulteration ofhoney is generally concerned by the consumers. Therefore, developing a newdetection method has become significant.In this study, the voltammetric electronic tongue is comprised of three electrodesstandard system coupled with a saturated calomel reference electrode, a gold workingelectrode and a platinum counter electrode. Voltammetric electronic tongue is basedon stand three electrodes sensors array, composed of electrochemical workstation andPC machines.In this thesis, an voltammetric electronic tongue was developed to classify threekinds of honey samples: six kinds of floral honey samples, five kinds of brand honeysamples and six different adulteation honey samples. The applied voltammetry iscomposed of three methods: cyclic voltammetry, different pulse voltammetry andsquare wave voltammetry. The database which is biulded by voltammetric electronictongue is analyzed by principal component analysis (PCA) to compare these threekinds of voltammetry. On this basis, fuzzy k-nearest neighbor (FKNN), fuzzyadaptive resonance theory (Fuzzy ARTMAP), radial basis function (RBF) are used topattern recognition. Results show that the recognition effect of Fuzzy KNN andFuzzy ARTMAP is better than that of RBF neural network. For adulteration honey,use partial least square to predict the adulterated concentration, and then can obtainthe fitting curve. The result shows that adulterated concentration prediction effectbased on the different pulse voltammetry is best. It is shown that voltammetryelectronic tongue can better distinguish the types of honey and the adulteratedconcentration of adulterated honey.

Related Dissertations
More Dissertations