he second method uses the BP neural

he second method uses the BP neural network's non-linear mapping and adaptive capabilities to simulate the complex non-linear relationship between input factors and output factors in actual yarn production, and uses key process factors and raw cotton quality factors as input variables The yarn quality prediction method based on the three-layer BP neural network and the yarn quality prediction method based on the four-layer BP neural network are designed respectively, and the prediction results of the two prediction methods are compared and analyzed.