Non-directional milk powder doping detection method based on flow discrimination model

Abstract

The invention discloses a milk powder doping non-directional detection method based on a flow discrimination model, which ingeniously converts milk powder doping information into a deep credible space boundary identification problem of a Raman hyperspectrum of normal milk powder, and can solve the non-directional detection problem of milk powder without depending on priori information of dopants. And only normal milk powder samples with recognized quality or strictly controlled quality need to be added for subsequent maintenance and upgrading. In model training, aiming at the problems of low migration efficiency, poor applicability and the like of a current deep learning model, the method designs a distribution transfer strategy to calculate high-dimensional complex Raman hyperspectral probability density, realizes direct numeralization control of a non-directional discrimination model training process, and greatly optimizes the model migration efficiency; according to the related characteristics, the wide applicability of the flow discrimination model is remarkably improved, the milk powder quality safety of China is powerfully defended, and the method is expected to be further expanded to non-directional detection of doping of other food systems.

Description

technical field [0001] The invention relates to the technical field of food safety detection, in particular to a Raman hyperspectral imaging flow discrimination model and method for non-directional detection of milk powder doping. Background technique [0002] As a high-quality nutritional source, milk powder products are an important part of the current human diet. However, milk powder products are faced with the threat of doping behavior driven by economic interests, and food safety incidents are emerging one after another, especially the milk powder doping incidents represented by melamine and big-headed milk powder, which not only caused serious harm to the health of infants and young children, but also severely hit domestic production. The overall reputation of the dairy industry has yet to fully recover. Therefore, the adulteration of milk powder must be strictly regulated. [0003] Milk powder adulteration detection is an important part of milk powder quality and sa...

AI Technical Summary

Problems solved by technology

[0006] In view of this, the present invention provides a non-directional detection method for milk powder doping based on a flow discrimination model, which effectively overcomes the problems of low efficiency, narrow coverage, high cost, and cumbersome operation of traditional directional detection methods, and is expected to be applied to massive milk powder authenticity regulation

Images