Quality analysis of wheat based on BP neural network and Near Infrared Reflectance Spectroscopy

With its quick, simple, nicety and nondestructive characteristic, NIRS (Near Infrared Reflectance Spectroscopy) is a new method for quality analysis of wheat. In this paper, a new method to model-building for wheat quality analysis with NIRS is presented. Opposite near infrared parameters shield the testing accuracy from outer disturb and random factors. Local minimization is escaped, and a high convergence velocity is reached by modified BP algorithm. The experimental results indicate that a high-accuracy testing results can be get in spite of large disturb from temperature and moisture.