苎麻纤维性能与成纱质量的人工神经网络分析

本文分别采用三种方法-BP神经网络、灰色关联分析结合BP神经网络、主成分结合BP神经网络根据苎麻纤维的性能建立了成纱性能的预测模型。采用灰色关联分析和主成分分析可以减少BP神经网络的输入节点数，提高预测结果的精度和稳定性。与单纯的BP神经网络的预测结果相比，灰色分析结合BP神经网络和主成分分析结合BP神经网络的预测结果更准确，在对成纱性能进行预测时，预测值与实测值之间的平均相对误差均明显下降。

Analysis of Ramie Fiber Quality by Artificial Properties and Yarn Neural Network

In this paper, three methods, pure BP neural network, grey relational analysis combined with BP neural network and principal component analysis combined with BP neural network were applied to build models of predicting yarn quality on the basis of ramie fiber properties. The last two methods were expected to reduce the input node numbers of BP neural network, and the network structure could be simplified, therefore the prediction accuracy and stability could be improved. Compared with pure BP neural network, the results gotten from the last two methods were both better, the mean relative error between the predicted results and the measured results of ramie yarn quality, such as the strength, strength irregularity, unevenness and neps, were all reduced greatly.