神经网络技术在织机生产状况预测中的应用

Application of artificial neural network to prediction of loom production

为提高织机生产效率,研究了基于优化神经网络的织机生产运转状况预测方法。针对BP网络模型的缺点,在反复实验的基础上对BP网络参数、算法进行改进,建立了织机生产运转状况预测模型,并与传统的BP神经网络预测方法进行比较。实验结果表明,利用改进的BP神经网络预测织机生产运转状况时,网络收敛速度快,预测精度高,优于传统的BP网络模型,能够取得较好的预测效果,从而准确设置织机生产参数,确保织机正常运转。

In order to forcast loom production operation,optimize the parameters of loom production,and then improve the efficiency,we studied the prediction of loom production based on the optimized neural network.In light of the disadvantages of traditional BP network the network parameters and algorithm is improved to make prediction model of loom production.Results show that improved BP neural network has high convergence speed and high forecast accuracy,helps accurately set production parameters and ensures normal operation of looms.