连续型贝叶斯网络在制丝工艺过程稳态优化中的应用  

Application of continuous Bayesian network to steady state optimization of tobacco primary processing

作　　者：周冰 唐军 文里梁 侯岩 陈爱明 何邦华 温亚东 曾仲大 ZHOU Bing;TANG Jun;WEN Liliang;HOU Yan;CHEN Aiming;HE Banghua;WEN Yadong;ZENG Zhongda(China Tobacco Yunnan Industrial Co.,Ltd.,Kunming 650231,China;Dalian ChemDataSolution Information Technology Co.,Ltd.,Dalian 116023,Liaoning,China)

机构地区：[1]云南中烟工业有限责任公司,昆明市650231 [2]大连达硕信息技术有限公司,辽宁省大连市116023

出　　处：《烟草科技》2019年第12期93-99,共7页Tobacco Science & Technology

基　　金：烟草行业卷烟工艺与装备研究重点实验室开放课题“制丝关键因素影响卷烟质量的复杂网络系统研究与应用”（2017GYSYS04）

摘　　要：为弥补传统制丝工艺过程稳态优化方法的不足,基于贝叶斯网络分析方法构建了以工艺参数为节点、影响关系为路径、影响大小为路径参数的网络模型,通过定量分析工艺参数间的关系,实现制丝过程中关键参数的优化调整。以昆明卷烟厂制丝线数据为对象对网络模型进行验证,结果表明:6个关键参数的网络模型预测值与真实值间的偏差均在1%以内,预测值与真实值间平均偏差为0.52%,优于真实值与工艺标准设定值间平均偏差(0.85%),表明该模型具有较高的准确性。该技术可为进一步优化和完善制丝工艺过程提供支持。In order to improve the existing steady state optimization methods for technological process in tobacco primary processing,a network model was established by taking technological parameters as network nodes,influential relationships as paths and influencing degrees as path parameters on the basis of Bayesian network analysis method.The key parameters in primary processing were optimized via quantitatively analyzing the relationships between technological parameters.The established network model was validated with the data of tobacco primary processing lines in Kunming Cigarette Factory,and the results showed that the deviations between the predicted values and real values of six key parameters were all less than 1%with the average deviation of 0.52%,which was lower than the average deviation(0.85%)set by technological standards.It was indicated that the established model featured higher accuracy.This method provides a support for the further optimization of technological process of primary processing.

关 键 词：卷烟生产 制丝过程 贝叶斯网络 工艺参数 稳态优化 

分 类 号：TS432[农业科学—烟草工业]