Measure And Analysis Of Algorithm Optimization For Near Infrared Spectrum Of Dairy Product Composition Content

Protein,fat and lactose are the nutrients that people need in their daily lives.Dairy is one of the raw materials that can provide these nutrients.As the quality of life becomes higher and higher,the detection of dairy products becomes more and more important.Near-infrared spectroscopy analysis can measure the concentration of major components in dairy products,which is quickly,efficiently and without damage,and suiting for real-time monitoring in the process of dairy product processing,purchasing of raw milk,users or sales sites.At present,the domestic market sales of high-precision dairy measuring instruments mainly rely on imports or agents,domestic high-precision dairy measuring instruments are rare.To explore the reason,which is mainly due to frequency doubling or co-absorption of molecular vibration near-infrared spectroscopy caused complex spectral background,useful information extraction and other difficulties.Therefore,in the near infrared spectroscopy,the preprocessing,modeling and predictive analysis of spectral data are very important,especially the analysis of spectral pretreatment.The principal component weight reset preprocessing method is a new method which combining histogram specification and Fuzzy Analytic Hierarchy Process(AHP)for two-dimensional data information enhancement.This paper studies the method and applies it to the near-infrared in the process of spectral preprocessing analysis,the studies of paper has the academic research significance and popularization and application value,whether it is from methodological research to algorithmic design level,or from programming to analytic application.The main research work includes the following aspects:Firstly,the basic methods of Near Infrared Spectroscopy were studied,the mechanism of near infrared spectroscopy was analyzed,and the method of spectroscopy in the process of near infrared spectroscopy was discussed.Researching chemometric methods and their applications in near infrared spectroscopy,near infrared spectroscopy wavelength screening methods,spectral data pretreatment methods and quantitative correction methods.Secondly,study of Acousto Optic Tunable Filters(AOTF),in the spectral rang from 1000 to 1700 nm of dairy products to measure the concentration of fat,protein and lactose and other major components of the concentration of near infrared spectroscopy.The module design and related hardware and software design of AOTF based dairy component near infrared diffuse reflectance spectroscopy system were completed.Once more,the pretreatment method of spectral data in near infrared spectroscopy was studied emphatically.Discussing and studying the preconditioning method of fusion histogram prescribing and Fuzzy Analytic Hierarchy Process(AHP)for weight resetting of main components of dairy products,and completing the corresponding algorithm design and programming.Finally,the quantitative correction in near infrared spectroscopy was studied.Using Partial Least Squares(PLS)method was used to establish a linear calibration model of dairy product concentration near infrared spectroscopy system.The regression coefficient and spectral residual ratio method are used for wavelength screening,and the data before and after wavelength optimization are modeled respectively,the results of the prediction are compared and analyzed to verify the feasibility of spectral denoising and filtering using the method of principal component weight reset.