Pattern Recognition System Of Rapid Detection Of Sesame Oil Flavoring Based On Electronic Nose And Algorithm Study

As a kind of food additive used in sesame oil, sesame oil flavoirng is harmful, andexcessive intake of it can cause vomiting, dizziness and other symptoms, making damages tothe consumer’s physical and mind. Therefore, the establishment of an accurate and rapid testof sesame oil flavoirng has important practical significance.It is clearly defined in the National Institute of Standards of sesame oil—GB8233-2008thatthe sesame should be the only raw mateiral of sesame oil and the quality test methods ofsesame oil also be designated, such as sensory judgment, colorimetry, spectroscopicmethodology, etc. While all the means mentioned above has its limitation, either toosubjective, too complicated or too expensive, and cannot make qualitative or quantitativerapid discirmination on the content of the sesame oil flavoring.The model of combining the data acquisition card and the host computer was adopted inthis thesis, and the pattern recognition system of sesame oil flavoirng rapid detection by theelectronic nose technology was established based on the Lab VIEW virtual instrument. Themain achievements in the paper are as follows:(1)The hardware system of electronic nose for the detection of sesame oil flavoirng wasset up. In this study, the gas sensors which be more sensitive to sesame oil flavoring wereselected, and the gas detection sensor array with appropirate sensor and size was established.Signal from the sensor array could be acquired and transferred to the PC by the dataacquisition, and the results of the pattern recognition could eventually be displayed in thescreen.(2)The pattern recognition system for the detection of sesame oil flavoirng wasestablished. In this study, the software system for the sesame oil flavoirng detection wascompleted on the Lab VIEW. Artificial Neural Network (BP-ANN) and Principal ComponentAnalysis (PCA) in C++were chosen as the appropirate algorithm. And the feasibility andaccuracy of the two algorithms were verified by simulation data.(3)Specific expeirmental program for the detection of sesame oil flavoring as well as thestandard database was designed. By repeating experiments, a large number of data wasavailable, and they were used for the training of the Artificial Neural Network (BP-ANN) andPrincipal Component Analysis (PCA) algorithm. By the way of standard data analyzing andalgorithm improving, the database was ultimately established. Conclusions was drawn from the expeirments that compared with the conventional testmethods, the method based on the electronic nose technology has higher accuracy, shorterdetection time and be more operable, and it achieved non-contact detection of sesame oilflavoirng and has practical development and application value.