Research On Fast And Non-destructive Examination Technology Of The Bottled Vinegar With Near Infrared Spectroscopy

Near infrared spectroscopy(NIRS) as an indirect analytical technique,has many advantages.it is used to analyze unknown samples by establishing a model with calibration samples in a few minutes, avoiding the tedious work of traditional analytical techniques. Currently, Vinegar detection solution is poured into cuvettes collecting spectral data and analyze. Although this more quickly than traditional chemical analysis methods, but has not yet reached a real fast non-destructive testing. Therefore, In this paper, the vinegar under the not opened the packaging environment,a series of experimental analysis has been conducted.Collected colorless and transparent glass bottles, green beer bottles, large, small beakers, each of them as a reference to collect spectra of bottled vinegar.There are some problems in the process of experiment, using near infrared spectrometer collected bottled vinegar spectral data, and did not collect the spectral information, after thinking and teacher guidance, the acquisition device has been made some improvements, the problem has been resolved.This article using near infrared spectroscopy were four vinegar and bottled vinegar collecting spectral data, comparing the difference between their spectral analysis of the reasons for these differences, explaining cause crests and troughs, and observe the glass material of the spectrum detectedusing chemometric methods to reduce or eliminate this effect. Improving the efficiency of detection, to achieve true non-destructive testing.Classification of the four kinds of bottled vinegar:raw spectra collected bottled vinegar the spectra preprocessing to reduce the dimension of principal component Mahalanobis distance discriminant method, qualitative discriminant model. The results show that the correct classification of the results of the calibration set and prediction set to100%. The results prove that near-infrared spectroscopy technology can be of different varieties of bottled vinegar qualitative discrimination.The article also quantitative analysis of volatile acids in the vinegar, total acidity, soluble solids three indicators, near-infrared spectra of these indicators quantitative prediction model using neural network, in order to avoid the effects of noise in the spectral data, the selection750-1870nm spectral data, the first of the original data to reduce the dimension of the principal component. Select the contribution rate of the larger three main ingredients modeling results show that the total acid, volatile acids, soluble solids of the true value andthe predictive value of the correlation coefficients were0.9318,0.9670,0.7370. Studies have shown that the rapid detection of vinegar Strength Index (total acid, volatile acids, soluble solids) using near infrared spectroscopy method is feasible. The research provides rapid non-destructive testing of guiding significance and reference value.