Analysis of element content in Bromus inermis based on near infrared spectroscopy technology

Owing to the rich nutritional value of Bromus inermis and the advantages of fast, efficient, and non-destructive of near-infrared spectroscopy (NIRS), this study used NIRS and modified partial least square (MPLS) to analyze the contents of K, Ca, P, Mg, Fe, Al, and Mn in 101 samples of Bromus inermis. The results show that the ratio of performance to deviation for validation (RPD) of K, P, Fe, and Al in the established prediction models were 2.60, 2.88, 4.60, and 2.68, respectively. These results indicate that the prediction accuracy was high and can that the approach can replace the traditional analysis method for K, P, Fe, and Al composition in Bromus inermis. The RPD values of Ca and Mg were 2.37 and 2.38, respectively, and the prediction accuracy was weak. Our results show that when the number of samples is large, the contents of these elements can be approximately estimated for sample screening. The RPD value of Mn was low (1.93), the prediction effect was poor, and the error between predicted and measured values was large.