Research Of Liquor Fingerprints Based On Infrared Spectrum

Fingerprint means make the test samples by appropriate processing, and then use certain methods of analysis to get the sample spectrum or a chromatogram, such as spectroscopy and chromatography. Spectrum contains the sample feature information of different features. Because of the characteristic information of spectrum is different, makes the spectrum as fingerprints with specificity and representativeness, therefore be called fingerprint. Fingerprint emphasizes the similarity within the same species group, rather than the specificity of the individual. Through the fingerprint data collection, and analysis of the spectrum information, to make evaluation for the difference or consistency between these different experimental samples. Due to the rapid development of modern science and technology, the concept of fingerprint has been combined with data processing technology, and it has extended to broader areas of science.Fingerprint is a kind of comprehensive appraisal method which based on the research of the sample components system. This technology can have accurate sampling and efficient identification, it can be well used to evaluate the authenticity and performance of the sample preparation quality. Fingerprint technology includes chromatography and spectroscopy, etc. Infrared spectroscopy is a kind of fingerprint technology which has a good accuracy and reliability, therefore this article uses the infrared spectroscopy for liquor related testing, aimed at to distinguish the flavor types between liquors and detection the alcohol concentration of liquor. The research of fingerprint required to pass through preparation analysis、comparing、evaluation and validation processes. This research uses the Fourier infrared spectrometer to collect the infrared spectrum data of different flavor and alcohol concentration liquor samples. And uses the wavelet transform as the main data preprocessing methods, to remove noise of the infrared spectra of data. Finally uses the PLS and SVM as the main analysis method, to quantitative analysis of the infrared spectrogram of test sample data which after the pretreatment, and extract the characteristic parameters for training, to established the quantitative evaluation model between sample and information. Finally compare the three methods are used to categorize and alcohol liquor flavor types optimal concentration prediction model.A little checking determined that there is only a small amount of research which is about liqueur fingerprint detection, and more of them is just a preliminary qualitative analysis or lack of substantial spectrum data post-processing at present. The qualitative and quantitative prediction model obtained satisfactory accuracy achieves the expected effect. Through the establishment of the model, just to identify on the infrared spectra of liquor samples for quantitative analysis and comparison, then it may be accurate prediction of its flavor and alcohol concentration, so this research has achieved the expected effect. This research effectively makes fingerprint technology used in food safety and quality detection, it has made up for the blank of this field, and promotes the development and application of the fingerprint technology to a certain extent, enable us to see its bright development future.