毛精纺中织造质量的BP神经网络预报技术

Forecasting Technique of BP Neural Network in Weaving Processing of Wool Worsted

为了建立毛精纺织造过程的实际预报模型,采用BP神经网络模型和企业所提供的数据,并给出了实验验证。实验结果表明织机效率精度可达95%以上;织疵预报平均精度可达90%,但部分精度低,故相关影响因子还需发掘并输入模型。

The researchful aim of this paper is to establish an effective and practicable forecast model in weaving processing based on BP neural network and the data sets collected from a wool worsted mill. The accuracy and effectivity of the model are verified. The experimental results indicated that the predicting precision of loom efficiency is up to 96＆ and the precision for the prediction of weaving faults is 90% in average, but the accuracy of a few results is low, thus, it is necessary to find the other important factors as the input parameters to improve the model performance.