Study On Standardization Of Mid Infrared Spectra Of Milk

Compared to traditional detection technologies,mid infrared spectroscopy analysis technology has a series of advantages such as accuracy,convenience,and non-destructive.These advantages make it stand out and have been vigorously developed.Now it has been widely used in food,medical,chemical and other industries.However,in practical applications,it has been found that there are many difficulties that need to be solved,such as the standardization of spectra.Due to factors such as product model,composition module,usage time,and surrounding environment,there are certain differences between different mid infrared spectroscopy instruments in different regions,resulting in the inability to directly compare the test results on one instrument with those on another.Spectral standardization can solve this problem without re modeling.This article combines traditional spectral standardization methods with the characteristics of mid infrared spectroscopy and proposes an improved spectral standardization method,which effectively improves the accuracy of the prediction model and applies this method to the milk analysis system.The main research content of this article is as follows:(1)Due to the difficulty in transferring the mid infrared spectral prediction model between the master and slave machines,in order to solve this problem,this study introduced the standard normal transformation(SNV)method on the basis of segmented direct standardization(PDS)and proposed a new SNV-PDS method.This method first performs SNV processing on the host data and slave data,and then standardizes the obtained data through PLS processing.Compared to the original method,the improved method has improved accuracy and efficiency,making it more suitable for the standardization process of infrared spectroscopy in milk.(2)Based on the characteristics of the infrared spectrum in milk,the orthogonal signal correction method(OSC)and segmented direct standardization method(PDS)are combined to eliminate data with low correlation in the master-slave spectrum through orthogonal signal correction,and then the segmented direct standardization method is used to transfer the processed spectrum.This method,abbreviated as OSC-PDS method,has better standardization performance compared to the original method.(3)By studying the spectral standardization process and analyzing the impact of baseline shift and offset phenomena on the standardization results,a new MSC-PDS method was proposed by combining multiple scattering correction method(MSC)and segmented direct standardization method(PDS).The experimental comparison results show that compared to the baseline method,this method has better standardization results and a smaller error rate.(4)This study also developed a milk analysis system based on the improved segmented direct standardization method and a universal mid infrared spectrometer,which integrates the improved method.This function covers various steps of milk standardization,including sample partitioning,spectral preprocessing,standardization parameter determination,model establishment,and various stages of standardization.This system can accurately and efficiently standardize the master-slave machine milk spectrum,and has high application value.