基于BP神经网络的织物折皱等级评定

Assessing Fabric Wrinkle Degree Based on BP Neural Network

采用改进的BP神经网络客观评定织物的折皱等级,网络结构为3层结构,输入样本数据经BP神经网络学习训练后,很快达到较小的误差收敛,从而确保评价的准确性和可靠性。

In this paper, the modified BP neural network is adapted to assess fabric wrinkle degree objectively. It has three layers in the network structure, the sample data is input BP neural network, then the data is learned and trained by BP neural network, it will attain convergences with less error quickly, which assures the accuracy and reliability of assessment.