Detection System And Modeling Of Milk Macromolecules

There are more and more questions about the safety of dairy products, the milk is an important source of dairy products, it is arising from the wider concerns to detect accurately about the content of the major compositions of milk, it is principal to detecte the content of fat and protein as macromolecules in the milk. The most common method is the chemical analysis method to detecte the content of fat and protein as macromolecules in the milk, however, the detect cycle of this method is too long, it is difficult for rapid detection, the physical methods can solve this problem, however, more mature, such as ultrasound and spectral analysis method, there are common problems that the stability of emission source and measurement accuracy are not good, in theory these methods should have to often calibrate those instruments, these methods can not detect accuracy in a long time.In the paper a new way to measure milk constituents is carried out. The new technology utilizes the scattered light to transmitted light ratio of laser light to token the amount of protein and fat in the milk powder. This method has solved the problem so that the stability of the emission source is good.at a certain scope, the fluctuations on the light intensity of Emission source can be effectively eliminated by the scattered light to transmitted light ratio of laser light, and achieve the purpose of detecting the milk constituents.The paper discusse and analyse this method, and then, designed a detection device to detect the content of fat and protein in the milk, research on the scattering and transmission characteristics of fat and protein in the milk, after that, we research on the relation between the milk protein and fat and the ratio of scattered light-to-transmitted light, so, designed the detect system of milk constituents by this method, this system mainly include power supply circuit, Sampling system, Light intensity detection, A/D conversion and display unit.Finally in the experimental process, taking into the influence of the temperature and pressure on the system. Use the fitting tool of the matlab, set up five models to the milk protein and fat: linear model poly2, linear model poly3, general model exponential, general model gauss and general model power, finally, compare to the model accuracy, we choose the linear model poly3 as the best relationship model between the concentration of milk constituents and the ratio of scattered light-to-transmitted light.