基于改进BP神经网络的气流烘丝机控制方法研究

Control Method of Air Drying Machine Based on Improved BP Neural Network

随着智能制造的发展和两化融合在烟草行业的持续推进,应用于制丝生产各关键环节的自动化和智能化控制方式也层出不穷,在提升加工均质化水平的同时,很大程度上减少了人员和成本的投入。基于改进BP学习算法的神经网络模型,针对制丝HDT气流烘丝机提出了一种基于预测的智能控制方法。方法将对HDT气流烘丝机的控制分为两个阶段,并以二次加料出料含水率、切丝含水率、环境温湿度、工艺流量等因素为输入,实现对冷床出口含水率的预测,并基于预测结果实现对工艺流量的自动调节。最后,通过测试,验证了模型的有效性。

Based on the neural network model of improved BP learning algorithm,a predictive intelligent control method for HDT air dryer is proposed in this paper.The control of HDT air drying machine is divided into two stages,and the water content of secondary feeding and discharge,water content of cutting,ambient temperature and humidity,process flow rate and other factors are taken as input to realize the prediction of water content of cold bed outlet,and the automatic adjustment of process flow rate is realized based on the prediction results.The validity of the model is verified by testing.