Dissertation > Excellent graduate degree dissertation topics show

Study on Rapidly Detecting the Quality of Milk by Near -Inrared Spectroscopy

Author: LiYanXiao
Tutor: DongYing
School: Jiangsu University
Course: Of Food Science
Keywords: Near Infrared Spectroscopy Milk Partial least squares model Wavelength filtering
CLC: TS251
Type: Master's thesis
Year: 2007
Downloads: 98
Quote: 0
Read: Download Dissertation

Abstract


Detection of conventional milk quality indicators often rely on conventional physical and chemical analysis, this method is not only cumbersome and time-consuming, is not suitable for online analysis. Near infrared spectroscopy technique has the advantages of rapid, non-destructive, non-polluting. In recent years, the milk, the main ingredient is detected using near infrared diffuse transmittance spectroscopy has been more and more attention. This thesis focuses on several key issues in the milk, the main component of near infrared prediction model, research, and hope to provide a new reference for near infrared detection of milk and other liquid food quality. The main research content and results of the subject: the nature and composition of the milk, and the propagation of light in the milk is a very complex situation, chose the diffuse transmission collecting milk near infrared spectroscopy, and designed a simple device for milk spectral acquisition . By partial least squares (PLS) compare pretreatment built the accuracy of the prediction model, filter out the best method of pretreatment, the main factor for of milk fat PLS model 5 after multiplicative scatter correction (MSC), the relevant coefficient values ??and forecast root mean square error RMSEP values ??of 0.997 and 0.180, respectively; using a differential processing protein PLS model coefficients and predicted standard deviation of 0.980 and 0.179, respectively; quadratic differential treatment lactose PLS model correlation coefficient and the predicted standard deviation to achieve 0.880 and 0.121. Basic research to get the best model - PLS model could predict the main component of the milk, but the model is built on top of the points for all of 1868 in the range of 3800 ~ 11000cm -1 wave so much wavenumber point is bound to also give model has brought a lot of redundant information, and cause a decline in the ability of the model to predict In addition, a large number of data points also brought a huge amount of calculation, which ultimately affect the efficiency of online testing. Application the wavelength interval screening method - interval partial least squares method (iPLS), this study forward / backward interval partial least squares method (FiPLS / BiPLS) model to simplify the main component of the milk, and make use of it from the wavenumber range screened subintervals, a total of 187 wavenumber points for milk fat modeling, the main factor of the resulting model number, RMSECV value and RMSEP values ??were 0.09408 and 0.04531, shows a better model than the full spectrum predictions. On this basis, this study also uses a genetic algorithm combined with partial least squares (GA-PLS) screening wave number of points, milk fat model modeling wavenumber points reduce from 1868 to 35, the main factor of the model number is reduced to months, compared with FiPLS screening method resulting model, its simplicity and predictive ability has improved. Simplified by this method at the same time the model of the protein and lactose, and the resulting model is simple and the accuracy is higher.

Related Dissertations

  1. Research on Melamineanalysis Method in Milk Powder and Establish the System of Rapid Determination,O657.3
  2. The Research on Immobilization of β-galactosidase and Production Technology of Low Lactose Milk,TS252.4
  3. Study on Processing of Germinated Soybean Accumulating Peptides and Its Soybean Milk,TS214.2
  4. Monitoring Leaf Nitrogen Status in Rice with Near Infrared Spectroscopy,S511
  5. Study on Analysis of Microflora and Predictive Growth Model of the Major Spoilage Microorganisms in Pasteurized Milk,TS252.1
  6. Application of HACCP for Pasteurized Milk for Student,TS252.2
  7. Application of Phytosterols and Its Mechanisms in Dairy Cows,S823
  8. Investigation of the Methods of Drug Using in Dairy Cow Farm in China,S858.23
  9. An Observation of the Concentration Changes of MTX in Maternal Milk and Serum in Partial Placenta Increta,R714.22
  10. Exercise, fresh milk and calcium for osteoporosis in older women Mongolian populations affected,R580
  11. Our contemporary ethical entrepreneur Construction,B82-05
  12. Set Up Ion Chromatography Method for Simultaneous Determination of Nitrate Nitrite and Thiocyanide in Dairy Food,O657.75
  13. Simultaneous Determination of Melamine and Cyanuric in Food by Liquid Chromatograghy Coupled to Tandem Mass Spectromety,O657.63
  14. Plant Design of Infant Formula Milk Powders with Output 20 Tons Per Shift,TS252.8
  15. Study on the Separation and Control of Spoilage Microorganisms in Soy-mung Bean Milk,TS214
  16. Study on the Stability and Processing HACCP System of Corn Milk,TS210.4
  17. Experimental Study on the Ethylene/air Diffusion Flame Based on Near-infrared Emission Spectrometry Measurement,TK16
  18. TRF detection method combines traditional method to evaluate the quality of raw milk,TS252.7
  19. Instrument-to-Instrument Calibration Transfer and Standardization of Spectra,O657.33
  20. Skimmed milk powder small and medium- peptide content with the functional activity research,TS252.7
  21. DSP FPGA -based near-infrared detector key technology research,TN791

CLC: > Industrial Technology > Light industry,handicrafts > Food Industry > Slaughtering and meat processing industries
© 2012 www.DissertationTopic.Net  Mobile