Research on moisture content measurement based on improved BP arithmetic

How to reduce the measurement error caused by linear regression is the key in measuring tea moisture content with microwave transmission techniques. In this paper, an improved BP algorithm, which combines the genetic algorithm, is given for training artificial neural network to get the range of weights and thresholds. The method makes BP algorithm avoid getting into infinitesimal locally and has the merits of high prediction precision and rapid convergence. The results show that the mean squared error is 0.0116, the mean absolute error is 0.0738, the mean relative error is 0.1182 and the certain coefficient is 0.9863 between the predicted value and the real one.