| With the rapid development of integrated circuit,rapid prototyping and computer technology,impedance detection technology has the characteristics of fast,portable and low cost,and has a wide range of applications in agricultural products,medicine,industry and other fields.Because the price of different kinds of edible vegetable oil varied greatly in the market,some illegal producers and operators mixed cheap edible oil into camellia oil,which led to a price war to reduce the quality of products,resulting in the tort incidents of a vile nature.Compared with foreign countries,due to the influence of starting time,research basis and test conditions,most of the impedance detection technology still stayed at the stage of test equipment,and few devices were used for fast,real-time and accurate on-site detection.Especially in the commercialization of oil impedance detection,there was still a big gap compared with foreign countries,and the related reports were relatively few.To solve this problem,a rapid impedance detection and analysis platform was designed in this paper,which included the design of detection cell and sensor structure,impedance data acquisition and optimization of test control parameters.Camellia oil mixed with different proportions of cheap edible oil was taken as the research object,and the optimized platform was used to carry out the experiment.The electrical characteristic parameters of camellia oil mixed with other oils were measured by impedance detection technology,and the mathematical model of impedance and fatty acid content was established.At the same time,the identification of camellia oil samples was analyzed by pattern recognition algorithm to verify the feasibility and stability of the platform.The main conclusions of this study were as follows:1)The hardware and software of the impedance detection and analysis platform was proposed.The principle of impedance detection was based on the self-balancing bridge method.The design parameters of the interdigital electrode structure were determined as follows:the number,the length,the width and the spacing of the interdigital electrode were 20,5 mm,0.20 mm and 0.16 mm,respectively.The detection module was controlled by a computer and AD5933 impedance measurement chip programmed by a C51 microcontroller,for which IIC communication protocol was used.And modularized programming idea was adopted for the program structure.The main function modules of software design included system parameter settings,real-time data acquisition and storage,etc.2)The control parameters of the impedance detection and analysis platform were simulated and analyzed.The parameters of the platform were simulated and analyzed by COMSOL software,and the parameters of excitation voltage,frequence-scanning range and detection dose of sample were optimized to be 2 V,3 000 Hz to 10 000 Hz and 100 ?L,respectively.3)The control parameters of the impedance detection and analysis platform were tested and analyzed.When the excitation voltage was 2 V,the coefficient of variation of detection impedance was the smallest(0.333).The coefficient of variation of the detection impedance of different frequence-scanning ranged from 0.008 to 0.193,and the corresponding value of the frequency band of 5000 Hz to 10000 Hz was the largest.The overall trend of the detection results of different dose of sample showed consistency,and the range of the impedance response value of the frequency band of 3000 Hz to 6000 Hz was larger than that of the frequency band 6000 Hz to 10000 Hz.The paired p-value of 100 ?L of paired t-test results was larger,ranging from 3.49.e-11 t0 3.61e-11.4)The identification performance of camellia oil adulteration test for impedance detection and analysis platform was studied.The difference of impedance curves between oleic acid and camellia oil mixed samples indicated that oleic acid contributed more to the impedance information of camellia oil,and the recognition rate of the calibration set and validation set constructed by oleic acid mixed samples reached more than 86.67%.The recognition rate of the set constructed by linoleic acid mixed samples reached more than 80.00%.And the corresponding values of rapeseed oil mixed samples were 97.22%and 96.67%.SG smoothing and successive projections algorithm combining BP artificial neural network(SG-SPA-BPANN)was the best model and can be selected as a fast model for adulteration of camellia oil. |
PDF Full Text Request