The Design Of Milk Composition Parameters Photoelectric Detection System And Uncertainty Analysis

Nowadays, people increase not only the consumption of milk, but also theattention to nutrition. According to the relevant requirements of the dairy industryin our country, companies must provide the analysis information of the maincomponents while product and process dairy products; Meanwhile, raw milk mustcarry on the relevant quality test prior to the acquisition, the test focus on the corematerial in milk, protein and fat, which are also the important index of milkquality. Due to the complexity of the milk composition detection, detectionmethods currently used are more or less insufficiencies, mainly reflected in cost,operation, maintenance, stability and precision, etc.This paper designed a photoelectric detection system of milk ingredientsparameters, which is mainly based on the theory of the rate between scatteredlight and transmitted light, by detecting the optical information, and ultimatelyreflects the content of protein and fat in milk. Based on STC89C58RD+andMicroprocessor as the hardware development platform, this paper designed thelight intensity detection module, analog to digital conversion module,microprocessor control module, a display unit module, etc, then built aphotoelectric detection system of milk ingredients parameters, this kind ofmodular design enhances the flexibility and scalability of the detection system.Meanwhile, key issues were discussed and analyzed during the detection, whichcan optimize circuit structure, reduce the equipment cost and improve thedetection accuracy. Then programmed the underlying driver of prototype terminal,using multiple linear regression analysis method to optimize the Correlationmodel between milk ingredients and the rate between scattered light andtransmitted light, which makes the testing data more reliable. Finally, this paper designed a feasible experimental scheme of prototype for a large amount ofexperimental verification; meanwhile, analyzed the uncertainty of the system indetail, intuitively showed in this paper the detection level of the designed system.The final results showed that the model established in this paper associatedwith a correlation coefficient of0.96, detection accuracy of the system is betterthan95%, and the system of measurement uncertainty is0.082, meet the designrequirements, can achieve a well online test about milk composition parameters,has a good market application prospect.