Prediction Of Raman Spectra and Adulteration Concentration of Wheat Flour Based on Neural Network Modelling

Abstract

In order to quickly detect the adulteration of flour containing wheat flour quality testing, the Raman spectrum of wheat flour as the object of study, based on a number of data-based neural network to identify and determine the concentration of wheat adulteration. Firstly, the neural network system in this study borrowed the Raman wave number and Raman intensity curve from "Non-contact Detection of Benzoyl Peroxide in Flour Based on Raman Spectroscopy" as the training and testing set. Next, we created the neural network, set the training parameters, trained the network and simulated the test. Finally, the supervised learning process resulted in error analysis and comparison of results, and the Raman spectra of wheat flour and adulteration concentration were plotted visually.