摘要
为解决卷烟制丝生产过程中现有SPC监控方法存在的问题,提出了基于SPC和BP神经网络的质量监控方法。首先在传统控制图的基础上,提出了适合在线监控的移动窗口式控制图,然后分别建立了用于控制图模式识别和质量缺陷原因诊断的两个神经网络模型,最后通过松散回潮工序中出口物料含水率的质量监控实例,证明了该质量监控方法的有效性。
In order to solve the problems of SPC monitoring method during the cigarette primary process,a quality control method based on SPC and BP neural network is proposed.Firstly,a moving window control charts is proposed based on the traditional control chart,which is suitable for online monitoring.Then two neural network models are established for control chart pattern recognition and fault diagnosis respectively Finally,an application example about the control of material moisture content during loosening and conditioning is presented,which verified the effectiveness of this quality control methods.
出处
《工业控制计算机》
2011年第12期65-66,68,共3页
Industrial Control Computer
关键词
神经网络
制丝
模式识别
在线质量控制
neural network
tobacco primary processing
pattern recognition
on-line quality control
