摘要
为解决多工序制造过程关键质量特性识别中存在的质量特性间具有多重相关性以及数据高维度、小样本等问题,采用偏最小二乘回归改进Ada-LASSO方法并融合状态空间思想和Bootstrap方法实现多工序过程关键质量特性识别,给出了基于IAda-LASSO的关键质量特性识别步驟,通过仿真实验和应用实例说明了IAda-LASSO方法与LASSO和Ada-LASSO方法在质量特性间不同相关度下识别的有效性.研究表明,IAda-LASSO方法对多工序过程有良好的关鍵质量特性识别能力,特别当质量特性间有较强相关性时显著优于其它两种方法.
To solve the problems of multiple correlations,high data dimensions,small samples existing in the key quality characteristics identification of multistage manufacturing process,the Ada-LASSO method is improved by Partial least squares regression,and integrates the state space idea and Bootstrap method are integrated to identify the key quality characteristics in multistage process.The steps of key quality characteristics identification based on the IAda-LASSO are given,the effectiveness of the IAda-LASSO,LASSO and Ada-LASSO in different correlation degree of quality characteristics is demonstrated by simulation and application example.The research shows that the IAda-LASSO method has a good ability to identify key quality characteristics in multistage processes,especially when there is a strong correlation between quality characteristics,which is significantly better than the other two methods.
作者
王宁
张帅
刘玉敏
杨剑锋
WangNing;Zhang Shuai;Liu Yumin;Yangjianfeng(School of Business,Zhengzhou University,Zhengzhou 450001,China)
出处
《系统工程学报》
CSCD
北大核心
2019年第6期855-866,共12页
Journal of Systems Engineering
基金
国家自然科学基金资助项目(71672182,71711540309,U1504703,U1604262,U1904211)
河南省教育厅人文社科重点资助项目(2016-ZD-054).