原棉性能指标与成纱质量的人工神经网络分析

Artificial Neural Network Analysis of the Raw Cotton Properties and Yarn Quality

采用灰色关联法分析了原棉的性能指标对成纱质量的影响，并结合BP人工神经网络对成纱条干、成纱强度和成纱强度不匀进行了预测。与单纯的BP神经网络的预测结果相比，灰色关联分析法结合BP神经网络预测结果更准确，预测值与实测值之间的平均相对误差较小。

In this paper, grey correlation method was applied to analyze the effect of raw cotton properties on yam quality and BP artificial neural network was combined to predict yam evenness, strength, and strength irregularity. Compared with the pure BP neural network predictions, the result obtained from combining grey correlation method and BP neural network prediction was more accurate. The average error between predicted value and actual value is much smaller.