欢迎访问机床与液压官方网站!

咨询热线:020-32385312 32385313 RSS EMAIL-ALERT
基于Logistic回归的数据富裕环境下制造过程质量动态监控
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

国家自然科学基金联合基金项目(U1904211);国家自然科学基金面上项目(71672182);国家社会科学基金(20BTJ059);河南省科技攻关项目(232102211040)


Manufacturing Process Quality Dynamic Monitoring Based on Logistic Regression Model in Data Rich Environment
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对数字化工厂“数据丰富,信息贫瘠”环境下制造车间生产过程存在异常信息利用不充分而造成监控效率低的问题,提出基于SPC技术和Logistic回归模型的制造过程质量监控方法。将关键工序中相关质量数据采集到MES系统,根据过程质量数据绘制 T2控制图,然后利用Logistic回归模型挖掘过程异常信息,并通过 T2和Logistic回归的联合优化实现数字化工厂制造过程质量监控的动态调整。以某薄膜晶体管液晶显示器等离子薄膜沉积生产工序为实际案例,验证了该制造过程质量监控方法的可行性和有效性。因此,在 T2控制图的基础上通过Logistic回归模型考虑过程异常信息能够提高制造过程质量监控的灵敏性和鲁棒性。

    Abstract:

    Aiming at the problem of the abnormal information collected from manufacturing process is difficult to use and analysis based on “data rich,information poor” environment of manufacturing workshop in the digital factory,a novelty process quality monitoring technique was proposed based on SPC and Logistic regression.T2 control chart was established by the related data from key stage in MES system.The Logistic regression model was applied to dig abnormal data,and the manufacturing process dynamic monitoring model was optimized by T2 and Logistic model jointly.Taking a plasma-enhanced chemical vapour deposition process of TFT-LCD manufacturing company as example,the feasibility and effectiveness of the process quality monitoring method were verified.The result shows that considering the abnormal information by using Logistic model based on T2 control chart can improve the sensitivity and robust of process quality monitoring.

    参考文献
    相似文献
    引证文献
引用本文

张帅,吕晨,杨剑锋.基于Logistic回归的数据富裕环境下制造过程质量动态监控[J].机床与液压,2023,51(11):104-108.
ZHANG Shuai, LYU Chen, YANG Jianfeng. Manufacturing Process Quality Dynamic Monitoring Based on Logistic Regression Model in Data Rich Environment[J]. Machine Tool & Hydraulics,2023,51(11):104-108

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2023-06-25
  • 出版日期: 2023-06-15