Wheat Sedimentation Value Determination Based on Near Infrared Spectroscopy

The value of sedimentation is a comprehensive index to assess the quality of wheat, but using near infrared spectroscopy to determinate wheat sedimentation value is rarely reported. In this paper, we explore the feasibility and effectiveness of wheat sedimentation value determination based on near infrared spectroscopy. We establish the PLS calibration regression model of wheat sedimentation value and BP neural network calibration model of wheat sedimentation value. The roots mean square error of prediction(RMSEP) and prediction correlation coefficient(r) are 0.73554、7.3352 and 0.49118、10.015 respectively. And complete with models of wheat protein content built in the same circumstances to analyze reasons of poor wheat sedimentation value forecasted.