Font Size: a A A

Preliminary Study On Optimization Of Milk Model Based On Near-Infrared Spectroscopy Analysis

Posted on:2006-11-28Degree:MasterType:Thesis
Country:ChinaCandidate:S WangFull Text:PDF
GTID:2121360212971523Subject:Measuring and Testing Technology and Instruments
Abstract/Summary:PDF Full Text Request
Near-infrared spectroscopy analysis technique is rapid, low cost, noninvasive, not destroying environment. It is not only suitable for laboratory analysis, but also for in-field fast and real-time on-line analysis. It has important significance to determine milk components accurately and rapidly. This paper is concentrated on the preliminary study on milk components analysis and milk model optimization based on Near-Infrared Spectroscopy Analysis, including combined spectroscopy, elimination of outliers, data pretreatment。The main research content of the dissertation involves:Firstly, the abstraction of interesting spectral signal is studied with multivariate regression method. The characteristics of different multivariate regression methods are compared. The result shows that The Root Mean of Square Error of Prediction (RMSEP) are 0.329, 0.185, 0.164 for fat by MLR,PCR,PLS1, respectively. So, PLS1 can achieve a robust model.Secondly, the chemical and physical characteristic of milk, combined with the concepts of tissue optics, is used to study the constituents of transmission spectroscopy and diffuse reflection spectroscopy. It is advanced that the model of combined spectroscopy is better than the model of transmission spectroscopy and diffuse reflection spectroscopy. The lower RMSEP for fat, protein can be achieved simultaneously, which is improved by 26.2%~37.8% and lower than 0.12.Thirdly, the combination of different pretreatment methods of milk spectral data are studied, and the optimized parameters of different pretreatment methods are achieved from experiments. The results show that the precision of model is improved. After combined pretreatment, the RMSEP of fat, protein model made up from transmission spectroscopy is improved by 33.5%, 17.2%, respectively. Further more, after combined pretreatment and outlier elimination, the RMSEP of fat, protein model made up from transmission spectroscopy is improved by 60.3%, 68%, respectively. And this result is better than the result of combined spectroscopy.The aim of this study was fundamental research on the measurement of milk constituents by the near-infrared spectroscopy technique. And it laid the foundation for the further investigation.
Keywords/Search Tags:Near-Infrared Spectroscopy, Spectral analysis Milk Components Analysis, Combined Spectroscopy, Multivariate regression, Pretreatment
PDF Full Text Request
Related items